Add files using upload-large-folder tool

.gitattributes

@@ -59,3 +59,7 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
 smartwilds-camera-traps/analysis/camera_operational_occurrence_arrivals.csv filter=lfs diff=lfs merge=lfs -text
+smartwilds/analysis/arrival_traces_v2/camera_machine_occurrence_arrivals.csv filter=lfs diff=lfs merge=lfs -text
+smartwilds/analysis/arrival_traces_v2/camera_image_capture_arrivals.csv filter=lfs diff=lfs merge=lfs -text
+smartwilds/analysis/arrival_traces_v2/camera_operational_occurrence_arrivals.csv filter=lfs diff=lfs merge=lfs -text
+smartwilds/analysis/arrival_traces_v2/drone_machine_detection_crop_arrivals.csv filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,251 @@
+---
+license: cc0-1.0
+language:
+- en
+pretty_name: SmartWilds Metadata (The Wilds 2025)
+task_categories:
+- object-detection
+- image-classification
+- audio-classification
+tags:
+- biology
+- ecology
+- biodiversity
+- multimodal
+- sensor-networks
+- camera-trap
+- bioacoustics
+- drone
+- citizen-science
+- darwin-core
+- humboldt-eco
+- fair2
+- smartwilds
+- the-wilds
+- ohio
+size_categories:
+- 100K<n<1M
+description: >-
+  The metadata of the SmartWilds multi-modal field sensing dataset (The Wilds,
+  Ohio, USA, summer 2025): Darwin Core Event and Occurrence tables for camera
+  traps, bioacoustics, drones, and citizen science on one Event hierarchy, with
+  corrected timestamps and coordinates, dataset cards, data dictionaries,
+  provenance records, media manifests, arrival traces, the supplemental
+  material, and the defect inventory. Raw images, audio, and video are not
+  included.
+
+fair2_compliance:
+  findable:
+    doi: ""  # pending
+    metadata_registry: ["DataCite"]
+  accessible:
+    open_access: true
+    authentication_required: false
+  interoperable:
+    standards: ["Darwin Core", "TDWG", "Humboldt Eco", "FAIR2"]
+  reusable:
+    license_clear: true
+    provenance_documented: true
+  ai_ready:
+    machine_readable: true
+    structured_annotations: true
+
+darwin_core:
+  event_coverage:
+    start_date: "2025-06-30"
+    end_date: "2025-09-08"
+    decimal_latitude: 39.83
+    decimal_longitude: -81.74
+    coordinate_uncertainty_meters: 5000
+    locality: "The Wilds, Cumberland, Ohio, USA"
+    habitat: "Reclaimed grassland and lakes of a conservation center on a former surface mine"
+  occurrence_info:
+    kingdom: "Animalia"
+    sampling_protocol: "Camera traps, Song Meter acoustic recorders, drone video surveys, and structured visitor photography, integrated into Darwin Core Event and Occurrence records"
+
+configs:
+- config_name: camera_trap_occurrences
+  data_files: smartwilds-camera-traps/data/occurrences.csv
+- config_name: camera_trap_events
+  data_files: smartwilds-camera-traps/data/events.csv
+- config_name: camera_trap_images
+  data_files: smartwilds-camera-traps/data/images.csv
+- config_name: bioacoustic_occurrences
+  data_files: smartwilds-bioacoustics/data/occurrences.csv
+- config_name: bioacoustic_events
+  data_files: smartwilds-bioacoustics/data/events.csv
+- config_name: bioacoustic_recordings
+  data_files: smartwilds-bioacoustics/data/recordings.csv
+- config_name: drone_occurrences
+  data_files: smartwilds-drones/data/occurrences.csv
+- config_name: drone_events
+  data_files: smartwilds-drones/data/events.csv
+- config_name: drone_detections
+  data_files: smartwilds-drones/data/detections.csv
+- config_name: citizen_science_occurrences
+  data_files: smartwilds-citizen-science/data/occurrences.csv
+- config_name: citizen_science_events
+  data_files: smartwilds-citizen-science/data/events.csv
+- config_name: all_events
+  data_files: smartwilds/data/events_all_modalities.csv
+- config_name: all_occurrences
+  data_files: smartwilds/data/occurrences_all_modalities.csv
+---
+
+# Dataset Card for SmartWilds Metadata
+
+![Sensor deployment across The Wilds](smartwilds/images/smartwilds/deployment_map.png)
+*Sensor deployment across The Wilds: camera traps with viewing angles, Song Meter recorders, citizen-science photo locations, and drone flight tracks. Satellite imagery © Google.*
+
+**SmartWilds** is a multi-modal sensing dataset recorded with camera traps,
+acoustic recorders, drones, and visitor photographs at [The Wilds](https://thewilds.columbuszoo.org/)
+(Cumberland, Ohio) in summer 2025. This dataset is its metadata: Darwin Core
+tables for all four modalities on one Event hierarchy, with corrected
+timestamps and coordinates so they can be joined in space and time, the
+dataset cards, data dictionaries, provenance records, and media manifests of
+each modality, the arrival traces used in the systems analysis, the
+supplemental material of the paper, and the record of every defect found and
+repaired. Raw images, audio, and video are not included; they will be released
+separately.
+
+## Dataset Details
+
+- **Curated by:** Jenna M. Kline, Anirudh Potlapally, Elizabeth G. Campolongo
+- **Homepage and code:** <https://github.com/Imageomics/smartwilds> (documentation: <https://imageomics.github.io/smartwilds/>)
+- **Paper:** SmartWilds: Integrating, Aligning, and Documenting Multi-Modal Field Sensing Data (under review)
+- **Release build:** 2026-10-22; camera and merged tables updated 2026-11-08
+- **Collection:** [SmartWilds](https://huggingface.co/collections/ABC-Center/smartwilds-6af3c0aac0a67864cbf6edb4), which also holds the 2025 pilot media and the pilot paper
+
+| Component | Events | Occurrences | Notes |
+|---|---:|---:|---|
+| `smartwilds-camera-traps/` | 19,785 | 72,775 | 235,778-image inventory with corrected timestamps; operational boxes with original SAM3 geometry and triage label provenance |
+| `smartwilds-bioacoustics/` | 2,382 | 9,960 | 2,371 WAV recordings, BirdNET Analyzer v2.4 |
+| `smartwilds-drones/` | 55 | 318,215 | 39 videos, 15 flight sessions; frame-indexed detections and ground-projected positions |
+| `smartwilds-citizen-science/` | 14 | 622 | 1,103 media files, expert-verified labels |
+| `smartwilds/` (study level) | 22,242 | 401,602 | Merged tables, text records, sensor configurations, Campaign-2 arrival traces |
+
+All tables share one Event hierarchy rooted at `TheWilds_2025`:
+
+```
+TheWilds_2025 (study, 2025-06-30/2025-09-08)
+├── SmartWilds_camera_traps_curated (Project)
+│   └── SensorDeployment
+│       └── Image
+└── Other modality Projects
+    └── SensorDeployment / flight session / observation session
+        └── AudioRecording | MediaEvent
+```
+
+### Supported Tasks and Applications
+
+This dataset supports cross-modal ecological analysis, data-quality research, and sensing-systems research. Computer-vision and audio tasks become possible when the raw media are released; the metadata here is what makes those releases usable.
+
+**🔗 Cross-Modal Analysis:**
+- Joining camera, acoustic, drone, and citizen-science records by site, campaign, and time on one Event hierarchy
+- Corroboration of focal-species activity across sensors; sensor marginal value by place and hour
+- Diel activity and species occurrence per modality from corrected timestamps
+
+**🧹 Data-Quality and Metadata Research:**
+- Timestamp reconstruction (18 methods, with confidence per image), telemetry mapping across three drone platforms, raw-inventory reconciliation
+- Label provenance kept distinct: triage-resolved, human-reviewed, forced nearest-prototype, human-edited, validated, unverified
+- A 22-item defect inventory with detection family, repair, and residual risk
+
+**⚙️ Sensing-Systems Research:**
+- Acquisition- and inference-layer arrival traces for workload characterization (burstiness, diurnal rates)
+- Synthesis of coordinated multi-modal workloads and hardware feasibility prediction
+- Blank-image filtering, storage, and edge provisioning from the full image inventory
+
+**🤖 Computer Vision and Audio Tasks (with the media releases):**
+- Object detection and species classification (camera traps, drones, visitor photos) with provenance-labeled boxes and human reference sets
+- Multi-object tracking on frame-indexed drone detections
+- Bird call classification and sound-event detection against BirdNET outputs
+
+## Layout
+
+Each component folder is a complete FAIR² dataset package:
+
+```
+<component>/
+├── README.md             # the component's dataset card
+├── data/                 # Darwin Core events.csv, occurrences.csv, per-file tables
+├── metadata/             # DATA_DICTIONARY.md, deployments, configs, provenance/
+├── analysis/             # arrival traces (study level), camera arrival statistics
+├── scripts/              # generator, georeferencing, release build
+├── media_manifest.csv    # one row per raw file: release path, size, include flag, privacy screening
+└── CHECKSUMS.sha256
+SUPPLEMENTAL.md           # the paper's supplemental material
+```
+
+Column definitions are in each component's `metadata/DATA_DICTIONARY.md`.
+The defect inventory is `smartwilds/metadata/DEFECT_INVENTORY.md`.
+
+## Using the data correctly
+
+- **Pick a verification floor before analysis.** Almost every Occurrence is a
+  machine label, and the status vocabulary differs by table: camera-trap rows
+  carry a `labelProvenance`; drone and acoustic rows are `unverified` or
+  `validated`. The drone labels come from BioCLIP 2 (TreeOfLife-200M) in
+  open-world Tree of Life mode; species-rank accuracy on human-annotated
+  subsets is 0.271 and class-rank accuracy is 0.845 (SUPPLEMENTAL.md, S9).
+- **Occurrences are detection crops, not individuals.**
+- **Use the corrected timestamps.** Camera clocks were wrong on most
+  deployments; every image carries its correction method and confidence.
+- **Filter drone coordinates on their status** (`georef_status` in
+  `detections.csv`).
+- **Join fixed sensors by site code and campaign**, drones and citizen science
+  by coordinates or date; see the [joining guide](https://imageomics.github.io/smartwilds/datasets/joining/).
+
+## Personal and sensitive information
+
+Media in which a person can be individually identified are removed from the
+media release; people too small to identify (a person box under 80 px) are
+retained. The screening evidence and human-review outcome are recorded per
+file in the media manifests and the camera image inventory. Coordinates fall
+within a public conservation area and are not generalized, except for
+citizen-science records, which carry a park-level point.
+
+## Licensing
+
+Data: [CC0 1.0](https://creativecommons.org/publicdomain/zero/1.0/). Scripts: MIT.
+
+## Citation
+
+```bibtex
+@misc{smartwilds2025_metadata,
+  author    = {Kline, Jenna M. and Potlapally, Anirudh and Campolongo, Elizabeth G.},
+  title     = {SmartWilds Metadata (The Wilds 2025)},
+  year      = {2026},
+  publisher = {Hugging Face},
+  url       = {https://huggingface.co/datasets/ABC-Center/smartwilds_metadata}
+}
+```
+
+Please also cite the paper:
+
+```bibtex
+@article{kline2026smartwilds,
+  author  = {Kline, Jenna M. and Potlapally, Anirudh and Campolongo, Elizabeth G.},
+  title   = {SmartWilds: Integrating, Aligning, and Documenting Multi-Modal Field Sensing Data},
+  year    = {2026},
+  note    = {Under review}
+}
+```
+
+This dataset card follows the [FAIR² Drones Data Standard](https://imageomics.github.io/fair_drones/)
+(Kline and Campolongo, 2026).
+
+## Acknowledgements
+
+This project is supported by the [Imageomics Institute](https://imageomics.org), which is funded by the US National Science Foundation's Harnessing the Data Revolution (HDR) program under [Award #2118240](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2118240) (Imageomics: A New Frontier of Biological Information Powered by Knowledge-Guided Machine Learning), and by the [AI Institute for Intelligent Cyberinfrastructure with Computational Learning in the Environment (ICICLE)](https://icicle.osu.edu/), which is funded by the US National Science Foundation under [Award #2112606](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2112606).
+
+This work is also supported by the AI and Biodiversity Change (ABC) Global Center. The ABC Global Center is funded by the US National Science Foundation under [Award No. 2330423](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2330423&HistoricalAwards=false) and the Natural Sciences and Engineering Research Council of Canada under [Award No. 585136](https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=782440). This dataset draws on research supported by the Social Sciences and Humanities Research Council.
+
+Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation, Natural Sciences and Engineering Research Council of Canada, or Social Sciences and Humanities Research Council.
+
+Special thanks to Dan Beetem and [The Wilds](https://www.thewilds.org/) in Ohio for providing access to deployment sites and supporting this wildlife monitoring research. We thank the Columbus Zoo and Aquarium for their support in facilitating this project. All data collection, including drone flights, was conducted under the supervision of the Director of Animal Management, with permission from The Wilds Animal Care and Use Committee.
+
+This work used the [Ohio Supercomputer Center](https://www.osc.edu/) (Ohio Supercomputer Center. 1987. Ohio Supercomputer Center. Columbus OH: Ohio Supercomputer Center. <https://ror.org/01apna436>).
+
+## Dataset Card Contact
+
+Jenna M. Kline ([kline.377@osu.edu](mailto:kline.377@osu.edu))

SUPPLEMENTAL.md

@@ -0,0 +1,885 @@
+# SmartWilds — supplemental material
+
+> **Camera-trap update (2026-11-08).** The legacy camera counts,
+> Event hierarchy, first-generation Tree-of-Life method statements, and
+> camera detection-arrival rows in this historical supplement do not describe
+> the current camera release. The canonical camera release has 72,775
+> original-SAM3-geometry Occurrences, 19,785 camera Events, a 235,778-image
+> inventory, and 70,466 Campaign-2 operational-inference arrivals. Its
+> combined multimodal table has 401,602 Occurrences and 22,242 Events. See
+> [`CAMERA_TRAP_RELEASE.md`](CAMERA_TRAP_RELEASE.md) for the
+> release-facing scope and provenance. The historical sections remain for
+> auditability and must not be used as current camera-release statistics.
+
+Supplemental material for the SmartWilds
+datasets.
+This contains the full data inventories details, the timestamp-repair methods,
+the standards crosswalk, the field dictionary, the arrival statistics, and the
+complete defect inventory.
+
+Last updated: 2026-11-10.
+Release: SmartWilds release build 2026-10-22, camera tables updated 2026-11-08.
+It supersedes the 2026-07-24 integration pass. Paths such as
+`processing_output/...` are relative to the project archive; the released files
+are in the per-modality datasets of the [SmartWilds collection](https://huggingface.co/collections/imageomics/smartwilds).
+
+---
+
+## S1. Extended glossary
+
+
+| Term | Source | Meaning in this release | What it is *not* |
+|:--|:--|:--|:--|
+| Event | Darwin Core | A sampling act with a date, location, protocol, and `parentEventID`. Five levels in use: `Project`, `SensorDeployment`, `AudioRecording`, `MediaEvent`, `ObservationSession`. | Not an arrival in a stream. One Event can contain 10⁵ arrivals. |
+| Occurrence | Darwin Core | One organism record at a place and time, linked by `eventID`. Usually one detection crop. | Not an individual animal. Adjacent drone frames yield many Occurrences of the same animal. |
+| `eventType` | Darwin Core | The level of the hierarchy a row occupies. | Not the modality; modality is carried by `parentEventID` and `modality_type`. |
+| `basisOfRecord` | Darwin Core | `MachineObservation` for every detector-derived row; `HumanObservation` for citizen-science and text records and for the 12 human-added drone starling boxes. | Not a quality statement. See `identificationVerificationStatus`. |
+| `identificationVerificationStatus` | Darwin Core | Vocabulary differs by table. Post-triage camera traps: `human-reviewed`, `machine-resolved`. Drones and bioacoustics: `unverified`, `validated`, plus `human-supported; limited-resolution` for the drone starling supplement. Citizen science and text: `verified by expert`. Counts in S9.3. | Not a confidence score. `machine-resolved` and `unverified` are both machine labels; a `validated` or `human-reviewed` record was reviewed by a person. |
+| `dataGeneralizations` | Darwin Core | States whether a coordinate is an observed position, a sensor position, a projected estimate, or a platform-telemetry fallback. | Not a precision estimate; see `coordinateUncertaintyInMeters`. |
+| `informationWithheld` | Darwin Core | Records where locality precision was reduced for sensitive taxa. | — |
+| Sampling effort | Humboldt ext. | Acquisition quantity actually expended, per modality: camera observation-days, recorder-hours, flight-minutes, visitor sessions. | Not comparable across modalities. Not a detection count. |
+| Arrival | systems | One item entering a processing stream. Two layers measured: *acquisition* (image capture, WAV file) and *detection* (an Occurrence). | Not a Darwin Core Event. |
+| Deployment | systems | One physical installation of one device at one site over one interval. | Not a device. One device yields several deployments across campaigns. |
+| Duty cycle | systems | Fraction of wall-clock time a device was *scheduled* to acquire. | Not uptime. A device can be on-schedule and dead. |
+| Detection | systems | A model output: a SAM3 box, a BirdNET hit. | Not an identification. |
+| Identification | Darwin Core | A taxonomic assignment with a stated verification status. | Not implied by a detection. |
+| `locationID` | Darwin Core | Stable identifier for a **fixed** sensor position or a drone session origin. Null for locality-only records. | Not a comparable spatial sampling unit across modalities. |
+| Coverage | systems | Distinct (date, hour) pairs containing ≥1 Occurrence, per modality. | Not area surveyed. Not detection probability. |
+| Campaign | this release | One field deployment interval. Three: C1 Jun 30–Jul 3, C2 Aug 4–28, C3 Sep 8. | Not a Darwin Core term. |
+
+---
+
+## S2. Camera-trap deployment inventory
+
+Four camera models, HFOV 55–120°. Configuration varied in capture mode,
+resolution, burst length, PIR sensitivity, and inter-trigger delay. Every one of
+these is a workload multiplier (S2.3).
+
+### S2.1 Campaign 1 — Jun 30 to Jul 3, 2025
+
+| Deployment | Site | Model | HFOV | Facing | Mode | Photo | Video | PIR | Images | Notes |
+|:--|:--|:--|--:|:--|:--|:--|:--|:--|--:|:--|
+| TW01-CT02 | Nomad Ridge E | GardePro T5NG | 56.5° | SE (135°) | Hybrid | 4 MP, 3-burst | 1080p 10 s MP4+audio | High | 7,811 | Hourly timelapse, info strip on |
+| TW02-CT03 | Nomad Ridge W | GardePro A3 | 63° | SW (225°) | Photo+Video | 8 MP, 3-series | 1080p 10 s AVI+audio | High | 39 | Early battery death; 55 files total on card |
+| TW03-CT01 | Pasture D Feeder | GardePro T5NG | 56.5° | S (180°) | — | — | — | — | 6,661 | Near salt lick; config not documented |
+| TW04-CT04 | Lake Trail | Trail Camera G600 | 120° | NW (315°) | Photo+Video | 8 MP, 5-burst | 1080p 10 s MP4+audio | High | 570 | Images inverted (mount orientation) |
+
+### S2.2 Campaign 2 — Aug 4 to Aug 28, 2025
+
+| Deployment | Site | Model | HFOV | Facing | Mode | Photo | Video | Images | Notes |
+|:--|:--|:--|--:|:--|:--|:--|:--|--:|:--|
+| TW01-CT02 | Nomad Ridge E | GardePro T5NG | 56.5° | SE (135°) | Photo only | 4 MP, 3-burst | — | 26,573 | Removed Aug 11, relocated to dock |
+| TW02-CT01 | Nomad Ridge W | GardePro T5NG | 56.5° | SW (225°) | Photo only | 4 MP, 3-burst | — | 61,217 | Clock error; battery died Aug 18–25; 4 card clock states |
+| TW04-CT03 | Lake Trail | GardePro A3 | 63° | NW (315°) | Photo+Video | 8 MP, 3-burst | 1080p 10 s | 984 | Reset Aug 7; no data after Aug 25 |
+| TW06-CT04 | Zebra Shelter | Trail Camera G600 | 120° | NW (315°) | Photo+Video | 8 MP, 5-burst | 1080p 10 s | 40,907 | Battery reported 100% at retrieval (unusual) |
+| TW06-CT05 | Zebra Shelter (in) | Reveal Pro 3.0 | 55° | N (0°) | Photo+Video | 4K | 720p 10 s | 84 | Late deploy Aug 25–28 |
+| TW07-CT05 | Zipline Tower | Reveal Pro 3.0 | 55° | N (0°) | Photo+Video | 4K | 720p 10 s | **withdrawn** | People-only media were not uploaded; metadata only retained; deletion status unknown |
+| TW09-CT02 | Dock (left) | GardePro T5NG | 56.5° | S (180°) | Photo only | 4 MP, 3-burst | — | 90,834 | Largest deployment (38.5% of all images); relocated Aug 18; battery gap Aug 18–25 |
+| TW09-CT05 | Dock (right) | Reveal Pro 3.0 | 55° | SE (135°) | Photo+Video | 4K | 720p 10 s | 9 | Detection issues |
+| TW11-CT06 | Fence / lake | Reveal Pro 3.0 | 55° | NE (45°) | Photo+Video | 4K | 1080p 15 s | 89 | PIR level 9, 20 s delay, 1 bar cell service |
+
+**Totals:** 235,778 JPG + 5,113 video clips, 260 GB. 13 deployments, 12 retained.
+
+### S2.3 Configuration as workload multiplier
+
+| Parameter | Range in deployment | Workload effect |
+|:--|:--|:--|
+| Burst length | 3 or 5 frames/trigger | Direct 3× or 5× multiplier on raw images per ecological event |
+| Capture mode | photo-only / photo+video / hybrid | Video adds a 10–15 s clip per trigger: large storage, frame-extraction compute |
+| Resolution | 4 MP / 8 MP / 4K | Bytes and per-image inference time |
+| PIR sensitivity | High, Level 9, motion-detect | Higher sensitivity → more triggers and more blanks |
+| Inter-trigger delay | 0 / 10 / 20 s | Longer delay suppresses repeat triggers, lowering effective arrival rate |
+| HFOV | 55°–120° | Wide-angle covers more area at lower pixel density per animal |
+| Cellular uplink | Reveal Pro 3.0 only | The only device in the deployment that could participate in near-real-time coordination |
+| Timelapse | TW01-CT02 C1 only | Adds a scheduled arrival stream independent of PIR |
+
+---
+
+## S3. Acoustic recorder inventory
+
+All units: Wildlife Acoustics Song Meter Micro 2, 48 kHz.
+10 deployments across 6 sites. Duty cycle, not deployment duration, determines
+volume: uncompressed 48 kHz stereo WAV ≈ 33 MB/minute.
+
+| Deployment | Site | GPS | Schedule | Duty cycle | Raw WAV | Events | GB |
+|:--|:--|:--|:--|--:|--:|--:|--:|
+| **C1** | | | | | | | |
+| TW05-SM01 | Lake Trail Gate O | 39.83537, −81.73842 | 5-min clips hourly (**config error** — intended 1 hr sunrise + sunset) | 8.3% | 144 | 144 | — |
+| TW06-SM03 | Zebra Shelter | 39.83518, −81.73087 | 5 min/hr, 24/7 | 8.3% | 77 | 77 | — |
+| TW07-SM02 | Zipline Tower | 39.83288, −81.73455 | 1 hr sunrise + 1 hr sunset | ~8% | 12 | 12 | — |
+| TW08-SM04 | Fence w/ vegetation | 39.82900, −81.74378 | 5 min/hr, 24/7 | 8.3% | 78 | 78 | — |
+| | | | | | **311** | **311** | **6.0** |
+| **C2** | | | | | | | |
+| TW02-SM04 | Nomad Ridge Shelter | 39.82957, −81.73295 | 5 min/hr | 8.3% | 659 | 581 | 26.1 |
+| TW05-SM06 | Lake Trail Gate O | 39.82866, −81.73758 | sunrise–sunset | ~50–60% | 100 | 96 | 16.6 |
+| TW06-SM05 | Zebra Shelter | 39.82870, −81.73756 | 15 min/hr | 25% | 508 | 508 | 21.8 |
+| TW07-SM02 | Zipline Tower | 39.82891, −81.74392 | sunrise–sunset | ~50–60% | 339 | 339 | 56.4 |
+| TW08-SM03 | Fence w/ vegetation | 39.82892, −81.74392 | sunrise–sunset | ~50–60% | 341 | 341 | 56.7 |
+| TW09-SM01 | Dock | 39.83418, −81.73783 | sunrise–sunset (deployed Aug 14) | ~50–60% | 195 | 195 | 32.4 |
+| | | | | | **2,142** | **2,060** | **210** |
+| **Total** | | | | | **2,453** | **2,371** | **216** |
+
+**Raw-WAV reconciliation.** Of the 2,453 WAV paths on disk, 2,371 are
+`AudioRecording` Events: 2,307 with BirdNET detections and 64 on which BirdNET
+ran but detected nothing (Events with no Occurrences). The other 82 are not
+recordings of their own: 78 are Campaign-1 files copied into the Campaign-2
+TW02-SM04 directory during transfer, and 4 TW05-SM06 files produced no BirdNET
+result (three 3.9–22.0 s fragments and one unreadable file). Every path is
+classified in `data/recordings.csv` of the bioacoustics dataset.
+
+Note that TW02-SM04 produced **more than twice the files** of TW07-SM02 but
+**less than half the bytes** — the clearest illustration that duty cycle and
+file count are independent axes.
+
+---
+
+## S4. Drone platform and session inventory
+
+| Platform | Campaigns | Video | Telemetry format | Sessions |
+|:--|:--|:--|:--|--:|
+| Parrot Anafi 4K | C1 (Jul 1, 3), C2 (Aug 11, 18, 25) | 4K MP4 | Parrot JSON / GUTMA DX | 10 (+3 telemetry-only) |
+| DJI (unspecified) | C2 (Aug 11) | 1080p MOV | CSV, GPS + battery, 10 Hz | 1 |
+| DJI Air 2 | C3 (Sep 8) | 4K MP4 | Airdata CSV | 4 |
+
+**39** raw videos and **39** `MediaEvent` rows; one MediaEvent (`VID_001`)
+has zero Occurrences because it was intentionally excluded after review.
+**15** video-bearing `SensorDeployment` flight-session Events. Three more
+Aug 25 Parrot directories hold telemetry but no video and are not sampling
+Events. 90 GB total.
+
+Three platforms produced three telemetry schemas with different field names,
+units (feet vs. metres for altitude), and gimbal-angle conventions. Fields
+consumed by the projection script, with their per-platform sources:
+
+- platform GPS — `latitude_y` / `longitude_y`
+- AGL height — `height_sonar(feet)` when available, else `altitude (m)` / `altitude_m`
+- heading — `gimbal_heading(degrees)` or `compass_heading(degrees)`
+- pitch — `gimbal_pitch(degrees)`, with a configurable fallback when near-level
+- FOV — approximate per-platform constant
+
+---
+
+## S5. Camera-trap timestamp repair, in full
+
+235,778 images; 0 unmapped; 0 flagged for review. 18 correction methods.
+
+### S5.1 Raw timestamp source
+
+| Source | Images | Share |
+|:--|--:|--:|
+| Filename | 193,096 | 81.9% |
+| EXIF | 42,500 | 18.0% |
+| OCR of burned-in bottom overlay | 179 | 0.08% |
+| Manual read of bottom overlay | 3 | 0.001% |
+
+### S5.2 Final confidence
+
+| Confidence | Images | Share | Meaning |
+|:--|--:|--:|:--|
+| high | 105,488 | 44.7% | Verified correct, or anchored to a logged service event |
+| medium | 130,290 | 55.3% | Anchored interpolation assuming a constant clock rate — **untestable without a second anchor** |
+
+### S5.3 All 18 correction methods
+
+| Method | Images | Pattern |
+|:--|--:|:--|
+| `aug_tw09_ct02_sd23_correct_after_move` | 59,568 | verified correct |
+| `aug_tw09_ct02_sd22_start_anchor` | 31,266 | per-card anchor |
+| `aug_tw06_ct04_sd02_start_anchor` | 27,734 | per-card anchor |
+| `aug_tw01_2025_clock_offset` | 26,558 | constant offset |
+| `aug_tw02_actual_august_clock` | 17,301 | verified correct |
+| `aug_tw02_sd01_sep_clock_anchor` | 17,262 | per-card anchor |
+| `aug_tw02_sd01_2023_clock_anchor` | 13,689 | constant offset (**clock reporting 2023**) |
+| `aug_tw06_ct04_initial_start_anchor` | 13,173 | per-card anchor |
+| `aug_tw02_sd11_sep_clock_anchor` | 12,965 | per-card anchor |
+| `june_tw01_correct` | 7,811 | verified correct |
+| `june_tw03_correct` | 6,661 | verified correct |
+| `aug_tw04_sd02_start_anchor` | 696 | per-card anchor |
+| `june_tw04_start_anchor_inverted_images` | 570 | per-card anchor + inverted mount |
+| `aug_tw04_sd03_start_anchor` | 288 | per-card anchor |
+| `bottom_overlay_ocr` | 179 | OCR of in-frame timestamp |
+| `june_tw02_start_anchor` | 39 | per-card anchor |
+| `aug_tw01_2023_tail_clock_offset` | 15 | constant offset (**2023 tail**) |
+| `manual_bottom_overlay_read` | 3 | human read |
+
+Only **14,472 images (6.1%)** were accepted as recorded
+(`june_tw01_correct` + `june_tw03_correct`). The remaining
+**221,306 (93.9%)** required reconstruction.
+
+### S5.4 Per-deployment repair summary
+
+| Dataset | Deployment | Location | Images | First corrected | Last corrected | Methods |
+|:--|:--|:--|--:|:--|:--|--:|
+| C1 | TW01-CT02 | Nomad Ridge E | 7,811 | 2025-06-30 12:18 | 2025-07-02 06:50 | 1 |
+| C1 | TW02-CT03 | Nomad Ridge W | 39 | 2025-06-30 12:34 | 2025-07-01 00:34 | 1 |
+| C1 | TW03-CT01 | Pasture D Feeder | 6,661 | 2025-06-30 13:06 | 2025-07-02 01:04 | 1 |
+| C1 | TW04-CT04 | Lake Trail | 570 | 2025-06-30 13:25 | 2025-07-03 16:56 | 1 |
+| C2 | TW01-CT02 | Nomad Ridge E | 26,573 | 2025-08-04 09:01 | 2025-08-10 10:15 | 2 |
+| C2 | TW02-CT01 | Nomad Ridge W | 61,217 | 2025-08-04 09:05 | 2025-08-18 09:07 | **4** |
+| C2 | TW04-CT03 | Lake Trail | 984 | 2025-08-07 10:38 | 2025-08-25 10:49 | 2 |
+| C2 | TW06-CT04 | Zebra Shelter | 40,907 | 2025-08-04 09:52 | 2025-08-26 15:35 | 2 |
+| C2 | TW06-CT05 | Zebra Shelter | 84 | 2025-08-25 15:09 | 2025-08-28 08:17 | 2 |
+| C2 | TW09-CT05 | Dock / Waterfront | 9 | 2025-08-11 10:18 | 2025-08-13 18:20 | 1 |
+| C2 | TW09-CT02 | Dock, then Nomad Ridge W after Aug 18 | 90,834 | 2025-08-11 10:27 | 2025-08-28 09:44 | 2 |
+| C2 | TW11-CT06 | Nomad Ridge W | 89 | 2025-08-18 09:00 | 2025-08-28 02:47 | 2 |
+
+Files: `processing_output/detections/drone_camera_analysis/camera_time_mapping/`
+— `camera_image_timestamp_map.csv` (per-image, 17 columns),
+`camera_timestamp_summary_by_camera.csv`,
+`camera_timestamp_summary_by_date.csv`,
+`camera_trap_timestamp_corrections.csv`,
+plus `.before_bottom_ocr` and `.before_ocr_interpolation` stages.
+
+---
+
+## S6. Drone ground projection
+
+**The campaigns were flown for population estimation and behavior, not
+geolocation.** Platform GPS localizes a detection to a known pasture, which is
+the operative spatial reference. Per-animal projection is an optional refinement
+provided for completeness; **no result in any of the three papers depends on
+it.** Monocular distance estimation from a moving platform to moving subjects is
+unreliable and is not established practice in aerial wildlife survey.
+
+| Status | Detections | Share |
+|:--|--:|--:|
+| `ok` (projected position available) | 247,340 | 77.9% |
+| `invalid_altitude` | 27,306 | 8.6% |
+| `missing_drone_gps` | 24,467 | 7.7% |
+| `horizon_ray` (camera near level) | 18,324 | 5.8% |
+| `out_of_range` | 122 | 0.04% |
+
+Method: convert the bbox bottom-center pixel to a camera ray, rotate by
+heading/pitch, intersect with a flat ground plane, project the offset from
+platform GPS. Dominant error sources: gimbal-pitch convention, terrain relief,
+altitude source, video stabilisation, FOV assumption, and whether bbox
+bottom-center is the animal's ground-contact point.
+
+Camera-trap coordinates use the pinhole method from
+<https://github.com/Imageomics/camera-trap-geolocation>: bbox height → range,
+bbox horizontal offset → bearing, projected from the camera GPS. Assumes flat
+ground and a uniform **0.9 m** target height — the dominant error term at a site
+holding animals from geese to Père David's deer.
+
+**Known analysis hazard.** An internal audit found a spatial figure that plotted
+projected animal positions and platform-telemetry positions together without
+distinguishing them: 15.66% of 292,106 plotted drone points were platform
+coordinates. The status field was present and correct; the script did not read
+it. The release carries that status on every drone Occurrence
+(`drone_georef_status` in `occurrenceRemarks`, `georef_status` in
+`detections.csv`); any analysis of drone coordinates must filter on it, and
+`examples/quickstart.py` shows the pattern. No automated check of downstream
+analyses ships with the release.
+
+---
+
+## S7. Standards crosswalk, in full
+
+| Required information | Standard | Status |
+|:--|:--|:--|
+| Taxon, rank hierarchy, vernacular name | Darwin Core | reused directly |
+| `eventDate`, `eventTime`, event hierarchy | Darwin Core | reused directly |
+| `decimalLatitude`/`Longitude`, `coordinateUncertaintyInMeters`, `geodeticDatum` | Darwin Core | reused directly |
+| `basisOfRecord`, `identificationVerificationStatus` | Darwin Core | reused as the automation boundary |
+| `dataGeneralizations`, `informationWithheld` | Darwin Core | reused for coordinate semantics and sensitive taxa |
+| `samplingProtocol`, `samplingEffort`, `sampleSizeValue`/`Unit` | Darwin Core + Humboldt | reused; effort reported per modality, never summed |
+| Inventory completeness, sampling scope | Humboldt | partially used |
+| Device model, sensor specification | SensorML concepts | re-expressed in platform tables (not SensorML-encoded) |
+| Derivation chain (raw → detection → identification) | PROV-DM concepts | re-expressed in remarks fields (not PROV-encoded) |
+| Platform, mission, annotation documentation | FAIR² Drones | dataset card template adopted |
+| Release principles | FAIR, CARE, TRUST | followed |
+| **Duty cycle and acquisition schedule** | — | **systems extension** (deployment `samplingProtocol` + config tables) |
+| **Trigger semantics** (PIR sensitivity, burst length, delay) | — | **systems extension** (config tables, S2.3) |
+| **Timestamp confidence and repair method** | — | **systems extension** (timestamp map, S5) |
+| **Per-detection coordinate provenance status** | — | **systems extension** (`drone_georef_status`, S6) |
+| **Typed missingness** | — | **systems extension** (S8) |
+| **Model identity, version, parameters, per-rank accuracy** | — | **systems extension** (S9) |
+| **Validator coverage and known gaps** | — | **systems extension** (S11) |
+
+The last six rows are the ones no standard we found expresses in a
+machine-readable way. We name them and use them consistently; we do not claim to
+have standardized them.
+
+---
+
+## S8. Missing data, broken down by type.
+
+
+
+| # | State | How it is computable | Example |
+|--:|:--|:--|:--|
+| 1 | Sensor active, nothing observed | Hourly bin inside a deployment window with no Occurrence | Most nocturnal camera bins |
+| 2 | Sensor inactive by schedule | Deployment-level schedule field | 5 min/hr recorders are off 91.7% of the time by design |
+| 3 | Sensor inactive by failure | Capture gap vs. logged deployment window, with cause | TW02-CT01 battery, Aug 18–25 |
+| 4 | Captured but not represented | Raw file inventory vs. Event table | 78 transfer-duplicate WAVs and 4 failed WAVs, each classified per path (S3) |
+| 5 | Metadata unusable for a purpose | Status/provenance column | 70,219 drone detections unprojectable (S6); until 2026-11-10, 41,524 drone Occurrences with a date but no time of day (S10, since recovered) |
+| 6 | Excluded in QC | Documented exclusion | People-only TW07-CT05 deployment excluded; one drone video reviewed and kept as a zero-Occurrence `MediaEvent` |
+
+States 1 and 2 have opposite implications for a coordination policy, which is
+why the schedule field is load-bearing and not documentation.
+
+---
+
+## S9. Models, parameters, and measured accuracy
+
+Manual annotation does not scale to sensor-network volume: human-reviewed
+labels are under 1% of the release and consumed the review capacity available.
+Semi-automated analysis is therefore the realistic condition, and we use
+current state-of-the-art models so that reported quality reflects what a
+practitioner adopting today's best tooling would obtain.
+
+Labels were produced in two generations, and the released tables carry both.
+The **first generation** labeled every SAM3 box with a BioCLIP Tree-of-Life
+taxon; it still describes the operational drone table and the merged
+cross-modal snapshot. The **second generation** (October–November 2026)
+replaced the camera-trap labels with embedding-based triage plus targeted
+human review (S9.1), and added a seven-video human-annotated drone reference
+set with a manual starling supplement (S9.2).
+
+| Stage | Model | Role |
+|:--|:--|:--|
+| Visual segmentation | SAM 3 | bounding boxes on camera-trap images and drone frames |
+| Visual classification, first generation | BioCLIP Tree-of-Life (TreeOfLife-200M) | top-1 taxon with full rank hierarchy, top-5 retained; operational drone labels |
+| Label triage, second generation | Resolve–Clear–Defer prototype triage over proposal embeddings | routes proposals to human review and resolves the rest by similarity to reviewed examples; operational camera-trap labels |
+| Acoustic detection | BirdNET | detections with frequency bounds and offset into the WAV |
+| Water/lake masking | SAM 3 | lake/water masks for habitat-association analysis |
+
+### S9.1 Camera-trap labels (post-triage)
+
+Operational camera-trap Occurrences are SAM3 animal proposals labeled by
+Resolve–Clear–Defer triage. The triage selects proposals for human review and
+resolves the remainder by similarity to reviewed examples; `labelProvenance`
+records the outcome on every row, with the triage round where recovered.
+
+| `labelProvenance` | `identificationVerificationStatus` | Occurrences |
+|:--|:--|--:|
+| `triage_reviewed` | `human-reviewed` | 1,760 |
+| `triage_resolved` | `machine-resolved` | 70,808 |
+| `triage_forced_1nn` | `machine-resolved` | 201 |
+| **Total** | | **72,769** |
+
+Forced nearest-neighbour labels did not meet the triage confidence rules and
+may be wrong; exclude them unless label noise is tolerable. Box geometry is
+never edited (`geometryProvenance = SAM3`), even when the label was human
+reviewed. The table keeps only reviewed or resolved animal proposals: cleared
+and background proposals, boxes with a shorter side below 32 px, boxes
+overlapping the per-resolution ROI mask, and frames on the privacy exclusion
+manifest are excluded, so **an image without an Occurrence is not evidence
+that no animal was present.** Twelve taxa occur: milu (Père David's) deer,
+Przewalski's horse, Canada goose, Sichuan takin, trumpeter swan, killdeer,
+European starling, barn swallow, and the broad labels bird, butterfly,
+dragonfly, and insect.
+
+A separate **human-reviewed reference set** of 2,067 boxes on 198 selected
+ungulate and waterfowl images (881 Canada goose, 627 milu deer, 349
+Przewalski's horse, 149 Sichuan takin, 61 trumpeter swan) is released beside,
+not inside, the Occurrence table. Taxonomy and geometry were both reviewed;
+1,552 boxes share a SAM3 proposal's geometry and 515 are human edits or
+additions, and the set is not filtered by the 32 px or ROI rules. The images
+were selected for review, so the set is not a random or exhaustive test split.
+Triage label accuracy against this reference set has not yet been reported.
+<!-- TODO(confirm-stats): add triage label accuracy vs. the 2,067-box reference set, or cite the Resolve–Clear–Defer paper's figures. -->
+
+### S9.2 Drone labels: operational table and seven-video reference set
+
+**Operational table.** The 318,203 machine drone Occurrences remain
+first-generation labels: a SAM3 box on every 15th frame with a top-1 BioCLIP
+Tree-of-Life taxon. The only measured accuracy for that labeler comes from
+human-annotated camera-trap subsets:
+
+| Analysis | Level | n | Precision | Recall | F1 / accuracy |
+|:--|:--|--:|--:|--:|--:|
+| SAM3 animal localization, camera traps | IoU ≥ 0.50 | 1,125 | 0.866 | 0.845 | 0.856 |
+| Tree-of-Life taxonomic assignment | class | 951 | | | 0.845 |
+| | order | 951 | | | 0.660 |
+| | family | 951 | | | 0.463 |
+| | genus | 951 | | | 0.322 |
+| | species | 951 | | | 0.271 |
+
+Usable statement: **class-rank labels support screening and triage; species-rank
+labels require curation; anything aggregating across ranks is uninterpretable.**
+A single headline accuracy anywhere in the 0.27–0.85 range would mislead.
+Corroborating evidence: the taxonomy crosswalk holds **12,577 label keys**, of
+which only **4** consist entirely of expert-verified records and 16 contain
+accepted names with mixed verification evidence.
+
+**Seven-video reference set.** Seven drone videos were human-annotated on
+1,089 sampled frames for four focal classes, with onager observations
+excluded: `20250703_parrotanafi_1`, `_2`, `_5` (C1), `20250811_DJI_01`,
+`20250818_parrot_01`, `_02`, and `20250825_parrot_01` (C2). The corrected
+reference holds **27,159** four-class boxes (the localization audit below was
+run against 27,180 before the final manual review): Père David's deer 21,177,
+Sichuan takin 3,293, Przewalski's horse 2,066, Canada goose 644. It is
+delivered as COCO JSON with provenance, a frame-to-proposal mapping, and the
+audit files. <!-- TODO(release): confirm the path of the seven-video reference set in the drone dataset. -->
+
+SAM3 localization against the reference (35,312 proposals, one-to-one
+matching):
+
+| IoU | Precision | Recall | F1 |
+|:--|--:|--:|--:|
+| 0.25 | 0.758 | 0.985 | 0.856 |
+| 0.50 | 0.749 | 0.973 | 0.847 |
+| 0.75 | 0.740 | 0.961 | 0.836 |
+
+Per-class recall at IoU 0.25: Przewalski's horse 0.961, Père David's deer
+0.985, Sichuan takin 0.993, Canada goose 1.000. Per-video recall at IoU 0.25
+ranges from 0.958 (`20250818_parrot_02`) to 0.999 (`20250825_parrot_01`).
+
+Resolve–Clear–Defer prototype triage was then run over the full proposal pool
+(seed 3, density-moderated k-center acquisition, calibrated rule k = 5,
+τ_fg = 4, τ_bg = 4). It spent 100 proposals on calibration and 1,610 on
+expert review (4.6% of the 35,212-proposal evaluation pool), handled the
+remaining 33,602 automatically, and got 28,040 of those right: 79.6% of the
+evaluation pool, or 83.5% of the automatic decisions. Zero proposals were left
+deferred.
+
+Four-species classification on the 25,897 matched, unreviewed foreground
+instances (78.4% Père David's deer, so always-deer scores 0.784 accuracy):
+
+| Method | Precision | Recall | F1 | Accuracy | Macro-F1 |
+|:--|--:|--:|--:|--:|--:|
+| Prototype triage, automatic decisions only | 0.913 | 0.850 | 0.880 | 0.850 | 0.789 |
+| BioCLIP 2 zero-shot, four prompts | 0.167 | 0.167 | 0.167 | 0.167 | 0.230 |
+| Tree-of-Life, open-world, exact focal-taxon match | 0.947 | 0.037 | 0.071 | 0.037 | 0.188 |
+
+Species-wise recall for triage: Père David's deer 0.918, Przewalski's horse
+0.687, Canada goose 0.590, Sichuan takin 0.551. The dominant automatic errors
+are wrong-species resolves into Père David's deer (takin 1,287, horse 595,
+goose 100) and 1,584 deer cleared as background. Read the table carefully:
+Tree-of-Life is evaluated open-world, so its 0.037 recall is exact top-1
+recognition of a focal taxon under an unconstrained vocabulary, not animal
+recognition accuracy; BioCLIP 2 zero-shot always answers one of four prompts
+and has no background option; triage abstentions count as no prediction. All
+figures are proposal-level repeated observations from seven videos, not
+independent animals.
+
+**Manual European starling supplement.** Twelve human-added *Sturnus vulgaris*
+boxes on one frame (`20250703_parrotanafi_2` / `VID_002`, frame 1320) were
+appended to the drone Occurrence table on 2026-11-01 with
+`basisOfRecord = HumanObservation`,
+`identificationVerificationStatus = human-supported; limited-resolution`, and
+remarks stating that no SAM3 proposal, embedding, or triage decision was
+involved.
+The identification rests on limited pixels in a single frame and should be
+interpreted cautiously. These records are excluded from the 27,159-box
+reference, the localization audit, the triage calibration and evaluation, and
+the species-classification comparison; none of the numbers above change if
+they are dropped. The 12 rows were appended twice (24 rows, 12 distinct
+`occurrenceID`s) in both the drone and merged tables (defect 21, repaired in the release build).
+
+### S9.3 Verification status distribution
+
+Counts are per table, because the vocabularies differ. The merged table below
+is the pre-update state (2026-11-01), before the camera rows were regenerated
+from the post-triage tables (defect 22, resolved 2026-11-08).
+
+Merged `occurrences_all_modalities.csv` as of 2026-11-01:
+
+| Modality | Status | Records |
+|:--|:--|--:|
+| Camera traps (first generation) | `unverified` | 223,203 |
+| | `validated` | 946 |
+| Bioacoustics | `unverified` | 9,960 |
+| Drones | `unverified` | 317,559 |
+| | `validated` | 644 |
+| | `human-supported; limited-resolution` | 24 (12 distinct records; defect 21) |
+| Citizen science | `verified by expert` | 622 |
+| Text data | `verified by expert` | 30 |
+| **Total** | | **552,988** |
+
+Post-triage camera-trap Occurrence table (the released camera dataset):
+
+| Status | Records | Share |
+|:--|--:|--:|
+| `machine-resolved` | 71,009 | 97.6% |
+| `human-reviewed` | 1,760 | 2.4% |
+
+`validated` = 946 camera-trap + 644 drone bird records from the curated
+goose/swan review set (added 2026-09-20). The 946 camera records belong to the
+first-generation camera table and are not carried into the post-triage
+Occurrence table; the 2,067-box reference set takes their role.
+`verified by expert` = all 622 citizen-science + 30 text-data records.
+
+A verification floor is one predicate: treat `unverified` and
+`machine-resolved` as machine labels; `human-reviewed`, `validated`, and
+`verified by expert` as person-reviewed; and `human-supported;
+limited-resolution` as person-reviewed but low confidence.
+
+### S9.4 Why this must be published
+
+1. A taxon string is not reproducible without the model that produced it. Two
+   versions of the same classifier yield different taxa from the same image.
+2. Published uncertainty is what makes labels **re-derivable**. Machine labels
+   are a snapshot of a model, not ground truth. A release that keeps source
+   media, detection geometry, model identity, and per-rank accuracy can be
+   re-labeled when better models arrive and the label sets compared
+   meaningfully. One shipping only final taxon strings cannot.
+3. Aggregate accuracy hides the rank-dependent failure mode above.
+4. Verification status must be per record, not per dataset — a README caveat
+   cannot be filtered on.
+
+---
+
+## S10. Sentinel values: `nan` as a date string (repaired)
+
+In snapshot `darwin_core_latest_20260724`, the 2025-08-11 drone session carried
+`eventDate` = the literal string `nan` (a dataframe export of a missing value)
+and an empty `eventTime`. The QC summary reported `missing_eventDate=0` because
+the check tested for null/empty and `"nan"` is neither.
+
+| eventID | Occurrences |
+|:--|--:|
+| `TheWilds_2025_drone_20250811_DJI_01_DJI_0002` | 12,522 |
+| `TheWilds_2025_drone_20250811_parrot_01_P5190525` | 12,430 |
+| `TheWilds_2025_drone_20250811_parrot_01_P5190526` | 11,996 |
+| `TheWilds_2025_drone_20250811_DJI_01_DJI_0001` | 4,535 |
+| `TheWilds_2025_drone_20250811_parrot_01_P5180524` | 41 |
+| `TheWilds_2025_drone_20250811_DJI_01`, `..._parrot_01` (`SensorDeployment`) | — |
+| **Total** | **41,524** (13.1% of drone Occurrences) |
+
+**Repair (release build 2026-10-22).** `eventDate` was restored to
+`2025-08-11` from the flight identifier and marked as identifier-derived.
+Drones flew on 6 dates; the old count of 7 included the string `nan`.
+
+**Time recovery (2026-11-10).** The processed telemetry had no clock, but the
+five video containers carry a creation time. Whether an aircraft writes UTC or
+local time into that field is airframe-specific, so it was settled against
+flights with independently known times: the Parrot Anafi writes UTC (its
+2025-08-18 video agrees with the telemetry file name to within 2 s), and the
+DJI Air 2 writes UTC (its 2025-09-08 video agrees with the telemetry's
+`datetime(utc)`). The 2025-08-11 Parrot videos therefore started at 09:12:43,
+09:14:19, and 09:19:55 local. The DJI Mavic 3 videos read 10:29:39 and
+10:31:32; read as UTC that would be 06:29, before anyone was on site, so this
+airframe wrote local time, which the telemetry export file name
+(`…_[10-29-13]`) and the field log also support. The field log places the
+Parrot flight at Stop 1 (Nomad Ridge, lake survey) and the DJI at the dock,
+Stop 5; the JotForm check-ins run 09:32 to 10:48. `eventTime` is now set for
+the 5 `MediaEvent`s and 2 flights (start/end intervals) and for all 41,524
+Occurrences (video start + `frame_index` / fps, to the second). The DJI
+convention is inferred rather than verified, so those times should be read as
+minute-level. The record is `metadata/provenance/drone_time_recovery_20250811.json`
+in the drone dataset.
+
+The same audit found impossible coordinates (`nan`, longitude ≈ 500) on seven
+drone Events, now replaced with a validated static video location or left
+blank, and a missing study-root Event `TheWilds_2025`, now added.
+
+---
+
+## S11. Validation: checks run, and gaps
+
+### S11.1 Original checks (snapshot `20260724`)
+
+- no missing event links (`eventID` resolves)
+- no missing taxon
+- no missing `eventDate` (presence only; this is how the `nan` dates passed)
+- no duplicate `occurrenceID`
+
+### S11.2 Current release checks
+
+The generator now parses every date, time, and coordinate pair and rejects
+sentinel values (`nan`, `None`, `NULL`, `-1`, `1970-01-01`) before export. The
+release audit checks, over every modality and merged table:
+
+- source-to-merged parity for Events and Occurrences;
+- a single-rooted Event hierarchy with no dangling `parentEventID`;
+- Occurrence-to-Event links and ancestry;
+- parsed dates and times, and WGS84-bounded coordinates;
+- every Occurrence date inside its parent Event's span and the study window;
+- no absolute cluster paths, and every `associatedMedia` resolving to a
+  released file in the dataset's media manifest;
+- agreement between the generator and the released tables.
+
+### S11.3 Gaps that remain
+
+| Gap | Consequence |
+|:--|:--|
+| No check that analyses constrain `drone_georef_status` | Mixed-semantics spatial figures are possible (S6) |
+| No semantic content check | People in imagery are found by a person detector plus human review, not by the validators |
+
+**A passing validator is evidence about the validator.** The original checks
+certified 552,964 records clean while 13.1% of one modality had no usable
+timestamp. Validator coverage should be reported, not assumed.
+
+---
+
+## S12. Darwin Core field dictionary
+
+`occurrences_all_modalities.csv` — 44 columns:
+
+`modality_type`, `occurrenceID`, `eventID`, `catalogNumber`, `scientificName`,
+`kingdom`, `phylum`, `class`, `order`, `family`, `genus`, `specificEpithet`,
+`taxonRank`, `vernacularName`, `occurrenceStatus`, `individualCount`,
+`lifeStage`, `sex`, `behavior`, `basisOfRecord`, `recordedBy`, `identifiedBy`,
+`identificationVerificationStatus`, `identificationRemarks`, `eventDate`,
+`eventTime`, `decimalLatitude`, `decimalLongitude`,
+`coordinateUncertaintyInMeters`, `geodeticDatum`, `locality`, `locationID`,
+`country`, `stateProvince`, `county`, `establishmentMeans`,
+`degreeOfEstablishment`, `occurrenceRemarks`, `associatedMedia`,
+`associatedReferences`, `datasetID`, `datasetName`, `informationWithheld`,
+`dataGeneralizations`
+
+`events_all_modalities.csv` — 21 columns:
+
+`eventID`, `parentEventID`, `eventType`, `eventDate`, `samplingProtocol`,
+`sampleSizeValue`, `sampleSizeUnit`, `locality`, `locationID`,
+`decimalLatitude`, `decimalLongitude`, `coordinateUncertaintyInMeters`,
+`geodeticDatum`, `country`, `stateProvince`, `county`, `datasetID`,
+`datasetName`, `recordedBy`, `fieldNotes`, `eventTime`
+
+### S12.1 Event hierarchy composition
+
+| `eventType` | Rows | Granularity |
+|:--|--:|:--|
+| `Project` | 6 | study root `TheWilds_2025` + one per modality |
+| `SensorDeployment` | 37 | 12 camera + 10 ARU + 15 drone flight session |
+| `AudioRecording` | 2,371 | one per processed WAV file (64 with no detections) |
+| `MediaEvent` | 39 | one per raw drone video (1 reviewed, no Occurrences) |
+| `ObservationSession` | 17 | 13 citizen science + 4 text data |
+| **Total** | **2,470** | |
+
+### S12.2 Occurrence composition
+
+| Modality | Events | Occurrences | Dates | Hourly bins | `locationID`s |
+|:--|--:|--:|--:|--:|--:|
+| Study root | 1 | — | | | |
+| Camera traps | 13 | 224,149 | 29 | 624 | 7 |
+| Bioacoustics | 2,382 | 9,960 | 29 | 650 | 6 |
+| Drones | 55 | 318,203 | 6 | 10* | 15 |
+| Citizen science | 14 | 622 | 13 | 28 | 0 |
+| Text data | 5 | 30 | 4 | 4 | 6 |
+| **Total** | **2,470** | **552,964** | | | |
+
+\* includes the 2025-08-11 session's bins after its times were recovered on
+2026-11-10 (S10); the integration pass had counted 8. Camera dates and bins use
+the corrected timestamps of the release build (439 curated records re-timed).
+
+Camera-trap figures here are the first-generation table in the merged
+snapshot. The released post-triage camera dataset has 19,762 Events (1
+`Project`, 12 `SensorDeployment`, 19,749 `Image`) and 72,769 Occurrences, and
+the drone table gained 12 manual starling records on 2026-11-01 (S9, defects
+21–22).
+
+---
+
+## S13. Arrival characterization, both layers
+
+Window: Campaign 2, 2025-08-04 to 2025-08-28, 600 hours — the only interval in
+which all four modalities were simultaneously active.
+
+Definitions. *Acquisition arrival* = one image capture (from
+`camera_image_timestamp_map.csv`, corrected timestamps) or one WAV recording
+(from `AudioRecording` Event times). *Detection arrival* = one Occurrence with
+a resolvable date and time. `B = (σ−μ)/(σ+μ) = (CV−1)/(CV+1)` over
+inter-arrival times: −1 periodic, 0 Poisson, →1 maximally bursty.
+
+| Stream | Arrivals | Rate/hr | CV | B | Dates |
+|:--|--:|--:|--:|--:|--:|
+| **Acquisition** | | | | | |
+| Camera image captures | 220,697 | 367.83 | 2.06 | +0.347 | 25 |
+| Acoustic WAV recordings | 2,060 | 3.43 | 1.23 | +0.103 | 25 |
+| **Detection** | | | | | |
+| Camera traps | 208,054 | 346.76 | 7.26 | +0.758 | 25 |
+| Bioacoustics | 8,865 | 14.78 | 3.21 | +0.525 | 25 |
+| Drones | 94,762 | 157.94 | 216.25 | +0.991 | 3 |
+| Citizen science | 242 | 0.40 | 5.30 | +0.682 | 10 |
+
+Detection streams count machine-derived Occurrences only; curated review
+records are excluded. Acoustic recordings include the 64 zero-detection WAVs.
+The drone row includes the 41,524 Occurrences of 2025-08-11 since their times
+were recovered (S10); before that recovery the stream had 53,238 Campaign-2
+arrivals at 88.73/hr, CV 229.61, B +0.991 over 2 dates. Every stream is released as a versioned trace (`analysis/arrival_traces_v2/`
+in the study-level dataset) with a source-record ID per arrival, inclusion rules,
+and input checksums; earlier unversioned arrival summaries are deprecated.
+Acquisition arrivals are not reconstructible for drones (per-frame times are not
+retained) or citizen science (one submission per session).
+
+### S13.1 Diurnal profiles (mean arrivals per hour-of-day over 25 days)
+
+| Hour | Cam capture | Cam detect | Acoustic WAV | Acoustic detect | CS detect |
+|--:|--:|--:|--:|--:|--:|
+| 00 | 290.1 | 76.5 | 1.8 | 3.4 | 0.0 |
+| 01 | 296.3 | 86.9 | 1.8 | 3.2 | 0.0 |
+| 02 | 317.4 | 184.1 | 1.8 | 3.3 | 0.0 |
+| 03 | 300.3 | 128.6 | 1.8 | 4.3 | 0.0 |
+| 04 | 304.0 | 211.9 | 1.8 | 3.3 | 0.0 |
+| 05 | 300.6 | 133.3 | 2.8 | 21.1 | 0.7 |
+| 06 | 317.9 | 519.2 | **5.2** | **41.6** | 0.0 |
+| 07 | 321.8 | **674.4** | 4.3 | 20.1 | 0.0 |
+| 08 | 341.3 | 503.2 | 4.3 | 18.3 | 0.0 |
+| 09 | 407.2 | 597.6 | 4.6 | 18.4 | 1.7 |
+| 10 | 454.4 | 416.5 | 4.3 | 18.6 | **2.7** |
+| 11 | 534.4 | 495.1 | 4.3 | 16.8 | 1.4 |
+| 12 | 589.0 | 507.6 | 4.3 | 16.2 | 0.2 |
+| 13 | **611.8** | 496.8 | 4.3 | 16.2 | 0.4 |
+| 14 | 527.6 | 301.1 | 4.3 | 15.8 | 0.0 |
+| 15 | 446.0 | 285.0 | 4.3 | 14.6 | 0.0 |
+| 16 | 385.3 | 193.5 | 4.3 | 16.5 | 0.0 |
+| 17 | 326.1 | 536.7 | 4.3 | 16.4 | 0.0 |
+| 18 | 297.8 | 499.9 | 4.3 | 17.6 | 0.0 |
+| 19 | 286.4 | 393.7 | **5.2** | 36.2 | 0.0 |
+| 20 | 292.8 | 494.8 | 3.0 | 18.7 | 0.0 |
+| 21 | 304.8 | 392.9 | 1.8 | 7.2 | 0.0 |
+| 22 | 288.4 | 138.9 | 1.8 | 3.0 | 1.6 |
+| 23 | 286.1 | 53.9 | 1.8 | 3.8 | 1.0 |
+
+Drone detections are zero in every hour but 09 (26,619 arrivals summed across
+the two flight dates).
+
+Diurnal swing, max/min: camera captures **2.1×**, camera detections **12.5×**,
+acoustic acquisition **2.9×**, acoustic detections **13.9×**. Detection profiles
+are sharper than the acquisition profiles feeding them in every modality where
+both are measurable.
+
+These values come from the versioned v2 arrival traces. They supersede the
+earlier unversioned arrival summaries, whose event definitions were not
+recorded.
+
+---
+
+## S13.2 Field servicing effort
+
+Derived from the Campaign-2 Jotform logs in
+`thewilds_20250804_20250828/jotforms/` (two camera-trap exports, two
+bioacoustic exports). Every row is one sensor servicing action with a `DateTime`
+stamp, site ID, device ID, SD-card ID, team member, and free-text note.
+
+| Trip | Actions | Window | Elapsed | Median gap between actions | People |
+|:--|--:|:--|--:|--:|--:|
+| 2025-08-04 | 8 | 08:58–10:41 | 103 min | 16.0 min | 3 |
+| 2025-08-07 | 6 | 08:59–10:05 | 66 min | 3.0 min | 1 |
+| 2025-08-11 | 9 | 09:32–10:48 | 76 min | 3.0 min | 2 |
+| **Total** | **23** | | **245 min (4.1 h)** | | |
+
+Breakdown of the 23 actions: 11 dropoff (deployment), 6 pickup (retrieval),
+5 check-in (status only, no SD or battery swap), 1 combined. By modality:
+12 acoustic, 11 camera.
+
+Additional operational fields captured per action:
+
+- **Access difficulty**, 1–3 scale — recorded on 15 of 23 actions
+  (rating 3: 9 actions; rating 1: 5; rating 2: 1)
+- **Battery status** — e.g. 98%, 95%, 60%, "Low", "100%"
+- **SD-card storage** — e.g. "7% full", "26% full", "5GB"
+- **Camera heading** — compass octant, used by the camera-trap georeferencing
+- **Relocation notes** — free text landmarks ("Zebra shelter", "Zip tower",
+  "Lake trail"), which is how the TW09-CT02 mid-deployment move was recovered
+- **Team member name and email**
+
+
+**Scope boundary:** this is measured elapsed field time for the three retained
+mid-campaign Jotform exports only. Campaign-1 field notes have no clock times,
+Aug. 28 retrieval submissions are absent from the retained exports, and no
+per-stage analyst-time log exists. The release therefore reports no total
+campaign field effort and no measured desk-processing hours.
+
+**Coverage limits.** These exports cover the three mid-campaign trips only.
+Campaign-1 field notes
+(`thewilds_20250630_20250703/field_notes/*.md`) are narrative and contain no
+clock times, and the 2025-08-28 retrieval visits do not appear in the exported
+forms. So 4.1 h is logged servicing time, not total field effort.
+
+**Not recorded anywhere:** analyst time per processing stage. Field visits were
+logged with timestamps; the timestamp-repair and taxonomy-review work was not.
+This asymmetry is requirement R7 in the paper.
+
+---
+
+## S14. Reproduction
+
+Darwin Core generation:
+
+The scripts live in the SmartWilds code repository (`scripts/`). Set
+`SMARTWILDS_ARCHIVE` to the root of the field archive before running them; see
+`scripts/README.md`.
+
+```bash
+python scripts/darwin_core/generate_darwin_core.py --force
+python scripts/arrival_traces/generate_arrival_traces_v2.py
+```
+
+Camera-trap georeferencing:
+
+```bash
+python scripts/georeferencing/geolocate_camera_trap_treeoflife_predictions.py \
+  --input-dir  <archive>/processing_output/detections/camera_traps/methods_paper/annotations/sam3/treeoflife_sam3/predictions \
+  --config-csv scripts/georeferencing/camera_geolocation_config.csv \
+  --output-dir <archive>/processing_output/detections/camera_traps/methods_paper/annotations/sam3/treeoflife_sam3/geolocation/geolocated_predictions \
+  --overwrite
+```
+
+Drone georeferencing:
+
+```bash
+python scripts/georeferencing/geolocate_drone_treeoflife_predictions.py \
+  --prediction-dir <SAM3 per-video prediction CSVs> \
+  --detection-dir  <archive>/processing_output/detections/drones \
+  --config-csv     scripts/georeferencing/drone_projection_config.csv \
+  --output-dir     <archive>/processing_output/detections/drones/geolocation/geolocated_predictions \
+  --overwrite
+```
+
+### S14.1 Release discipline
+
+Integration passes are dated and immutable. Every mutation retains a
+before-copy and emits a JSON audit record. Passes differ **semantically**,
+not just additively — different visual-label semantics, corrected timestamps —
+so comparing numbers across passes is invalid in a way that is not visible in
+the numbers. **Any figure or table must name the release build it was computed
+from.**
+
+The 2026-10-22 release build supersedes the 2026-07-24 pass; each of its
+repairs (S10, S3, S4, S16) has its own JSON audit record, shipped under
+`metadata/provenance/`. The 2026-09-20 curated-bird update is reconstructible
+from:
+
+- `curated_bird_dwc_update_audit_20260920_015706.json`
+- `curated_bird_event_sync_audit_20260920_020046.json`
+- `curated_bird_project_sample_size_fix_20260920_020401.json`
+
+---
+
+## S15. Release logistics
+
+Each modality is published as its own Hugging Face dataset with its own DOI,
+and a study-level dataset holds the merged Event table, the text records, the
+arrival traces, and this supplement. The [SmartWilds collection](https://huggingface.co/collections/imageomics/smartwilds) joins all five.
+DOIs, the license, and the maintenance commitment will be listed here at
+publication. Contact: Jenna M. Kline (kline.377@osu.edu).
+
+## S16. Defect inventory, in full
+
+Every defect class found while integrating the SmartWilds campaigns (release
+build 2026-10-22; camera tables updated 2026-11-08). **Detected by** is
+what made the defect visible, grouped into five families: an external record
+(service log, field note, platform documentation), a redundant encoding inside
+the data, a cross-check between two artifacts, human inspection, or nothing.
+Rows 19-20 had no second record and could be bounded but not repaired.
+Rows 21-22 were found after the 2026-11-01 annotation update and were
+repaired in the release build (row 21) and the 2026-11-08 camera update (row 22).
+
+| # | Defect | Mod. | Scope | Detected by (family) | Repair | Residual risk |
+|---:|---|---|---|---|---|---|
+| 1 | Device clock wrong or unset | Cam | 221,306 / 235,778 images (93.9%) | Service log deploy/retrieve times (External record) | 18 per-deployment reconstruction methods (paper §timestamp reconstruction; S5) | 130,290 images (55.3%) only medium confidence |
+| 2 | Clock reset to factory epoch | Cam | 2 deployments, 1 reporting year 2023 | Service log; implausible date on its face (External record) | Anchored offset from known deploy time | Offset assumed constant over deployment |
+| 3 | Media split across SD cards with inconsistent clocks | Cam | TW02-CT01 (4 clock states), TW09-CT02 (2), TW06-CT04 (2), TW04-CT03 (2) | Card-level directory structure + service log (External record) | Per-card anchors reconciled against logs | Within-card drift unmodeled |
+| 4 | Timestamp only in burned-in overlay | Cam | 182 images (TW06-CT05, TW09-CT05, TW11-CT06) | The image itself (redundant in-frame encoding) (Redundant encoding in data) | OCR of bottom overlay; 3 read manually | OCR errors on 179 automated reads |
+| 5 | Camera mounted inverted | Cam | TW04-CT04, 570 images | Human review of imagery (Human inspection) | Flagged; images rotated downstream | --- |
+| 6 | Camera relocated mid-deployment | Cam | TW09-CT02 (dock → Nomad Ridge W, Aug 18) | Field note recording the move (External record) | Split into two logical location intervals | Single `locationID` spans both |
+| 7 | Battery exhaustion | Cam | TW02-CT01 Aug 18--25; TW02-CT03 C1 (55 files) | Capture gap vs. logged deployment window (Cross-check) | Recorded as a failure gap with cause | Lost observation window unrecoverable |
+| 8 | Non-wildlife content (visitors) | Cam | TW07-CT05, whole deployment | Human review of imagery (Human inspection) | People-only imagery not uploaded; deployment excluded from release | Card disposition after field not documented |
+| 9 | Recorder ran a schedule other than configured | ARU | TW05-SM01 C1 (144 files) | File inventory vs. intended config (Cross-check) | Actual schedule inferred and documented | Intended vs. actual effort differ |
+| 10 | WAV files with no event record | ARU | 146 / 2,453 paths (6.0%) | File inventory vs. event table (Cross-check) | Classified per path: 64 zero-detection recordings added as Events; 78 transfer duplicates and 4 failed files excluded | 4 fragments/unreadable files unrecoverable |
+| 11 | Heterogeneous telemetry schemas | Drone | 3 platforms, 3 formats, feet vs. metres | Platform documentation (External record) | Per-platform field mapping with unit conversion | Mapping assumptions per platform |
+| 12 | Drone video not processed | Drone | 1 / 39 raw videos | File inventory vs. event table (Cross-check) | Reviewed (no usable content); retained as a zero-Occurrence `MediaEvent` | --- |
+| 13 | Media references to purged scratch workspace | Drone | 16,069 frame paths, 307,260 Occ. | Resolving every `associatedMedia` path (Cross-check) | Replaced with the retained original video, flagged as a video-level fallback | Frame-level media link lost for those rows |
+| 14 | Unreviewed fine-rank taxon labels | Cam, Drone | Merged snapshot: 550,722 Occ. `unverified` (99.6%); post-triage camera table: 71,009 Occ. `machine-resolved` (97.6%) | Human-annotated validation subsets (Human inspection) | Camera labels replaced by embedding-based triage with human review and a 2,067-box reference set (S9.1); drone labels kept `unverified`, with a seven-video human-annotated reference set (S9.2); accuracy reported per rank and per method | Drone species-rank labels include implausible taxa; triage label accuracy against the camera reference set not yet reported |
+| 15 | Null written as the string `nan` | Drone | 41,524 / 318,203 Occ. (13.1%); 5 Events + 2 deployments | Date redundantly encoded in the `eventID`; time from the video containers cross-checked against the field log (Redundant encoding in data; External record) | Date restored from the identifier and flagged as derived (release build 2026-10-22); time of day recovered on 2026-11-10 from the video container creation times, with each aircraft's clock convention settled against reference flights and the field log and JotForm check-ins of the day (`provenance/drone_time_recovery_20250811.json`) | DJI clock convention is inferred from the field schedule, not verified on a reference flight; treat those times as minute-level |
+| 16 | Impossible drone coordinates (`nan`, longitude ≈500) | Drone | 7 Events | Domain check on coordinate pairs (Cross-check) | Valid static video location if inside study area, else blank | Those Events lack a flight-level position |
+| 17 | Curated records bypassed timestamp correction | Cam | 439 / 946 curated Occ. | Occurrence date outside parent deployment interval (Cross-check) | Re-timed from the corrected camera timestamp map (release build 2026-10-22) | --- |
+| 18 | Root Event row absent from Event table | All | 1 dangling `parentEventID` | Referential check over the Event table (Cross-check) | Study root Event added; parent links validated | --- |
+| 19 | Within-card clock drift | Cam | Up to 130,290 images | *none* (None (recorded once)) | Constant clock rate assumed | Unquantifiable timestamp error |
+| 20 | True deployment / retrieval times | Cam | Deployment metadata known to contain errors | *none beyond the erroneous records* (None (recorded once)) | Corrected occurrence dates used instead | True camera-days not reportable |
+| 21 | Manual starling supplement appended twice | Drone | 24 rows / 12 distinct `occurrenceID`s in the drone and merged Occurrence tables (2026-11-01) | Duplicate `occurrenceID` check (Cross-check) | Repaired in the release build: the 12 exact duplicate rows are dropped, leaving every `occurrenceID` unique (318,215 drone Occurrences) | --- |
+| 22 | Merged snapshot and camera release disagree on camera labels | Cam | 224,149 first-generation vs 72,769 post-triage Occ.; 13 vs 19,762 camera Events | Row-count parity between the camera release tables and the merged table (Cross-check) | Not yet repaired; regenerate the merged snapshot and the camera detection-layer arrival traces from the post-triage tables | Cross-modal totals (552,964) and camera detection-layer traces describe the first-generation labels |

smartwilds-bioacoustics/CHECKSUMS.sha256

@@ -0,0 +1,11 @@
+d86e7f58cc3c67f7b55b6d109b044ff3b2235eb165c19bb087edebe34ea2b32a  data/events.csv
+aa2d08cc1e94004295a6912ea853ed8dc0cc7ab913ca11826fcd2df4ba96c679  data/occurrences.csv
+67ab601613c43f416bcaae134a92d5f20eda8600c40c2905e6ecb4ec6c8afc38  data/recordings.csv
+65ae30dcfeaeeaceb547aaa59c9e78edd18f6879a05dbc7a4bd04ed0f2d90a95  media_manifest.csv
+572cc32cd74a99399864581f83a1db0b3747eba7ed40176027b1e7f6b923ff4e  metadata/DATA_DICTIONARY.md
+5337794469dacd255ff2e19c13e0693089d69add3bb3c7ae9db183ad0773404d  metadata/birdnet_settings.csv
+0d588f548dcb81e9856194fdbcd369e32d86cc6509b80b2cb0662521e8cf4ed2  metadata/deployments.csv
+54f47cbccd361b3ca35d07595ce8941a890ed5cb40dc087176e11070e43c8c96  metadata/provenance/RAW_WAV_RECONCILIATION_20261008.md
+e5b1a9730a77adf34d781b26682bac741b90019b02385a86163ba7cdf59b07fb  metadata/provenance/bioacoustic_event_repair_20261008.json
+83cba7ad8d340c28ac25519c9bf7845e034dfcdd6ef16e8a62fedd3f8a55293d  scripts/build_hf_release.py
+05ac332303b80a45edd9d9fe9f2927af25ebe126d9580c7aadb451cc0602068b  scripts/reconcile_bioacoustic_raw_wavs.py

smartwilds-bioacoustics/README.md

@@ -0,0 +1,238 @@
+---
+license: cc0-1.0
+language:
+- en
+pretty_name: SmartWilds Bioacoustics (The Wilds 2025)
+task_categories:
+- audio-classification
+tags:
+- biology
+- ecology
+- biodiversity
+- bioacoustics
+- passive-acoustic-monitoring
+- birds
+- birdnet
+- darwin-core
+- humboldt-eco
+- fair2
+- smartwilds
+- the-wilds
+- ohio
+size_categories:
+- 1K<n<10K
+description: >-
+  2,371 WAV recordings (215 GB) from 10 Song Meter Micro 2 deployments at six
+  sites at The Wilds (Ohio, USA), June-August 2025, with per-recorder duty
+  cycles, a one-row-per-file reconciliation table, and 9,960 Darwin Core
+  Occurrences (93,509 BirdNET detections across 128 bird species).
+
+fair2_compliance:
+  findable:
+    doi: ""  # pending the raw data release
+    metadata_registry: ["DataCite", "GBIF"]
+  accessible:
+    open_access: true
+    authentication_required: false
+  interoperable:
+    standards: ["Darwin Core", "Humboldt Eco", "FAIR2"]
+  reusable:
+    license_clear: true
+    provenance_documented: true
+  ai_ready:
+    machine_readable: true
+    structured_annotations: true
+
+darwin_core:
+  event_coverage:
+    start_date: "2025-06-30"
+    end_date: "2025-08-28"
+    decimal_latitude: 39.83
+    decimal_longitude: -81.74
+    coordinate_uncertainty_meters: 5000
+    locality: "The Wilds, Cumberland, Muskingum County, Ohio, USA"
+  occurrence_info:
+    basis_of_record: "MachineObservation"
+    taxa_included: "Aves (128 species)"
+    sampling_protocol: "Scheduled passive acoustic recording; BirdNET Analyzer v2.4 classification"
+
+sensors:
+  - type: "Autonomous recording unit"
+    manufacturer: "Wildlife Acoustics"
+    model: "Song Meter Micro 2"
+
+configs:
+- config_name: occurrences
+  data_files: data/occurrences.csv
+- config_name: events
+  data_files: data/events.csv
+- config_name: recordings
+  data_files: data/recordings.csv
+---
+
+# Dataset Card for SmartWilds Bioacoustics
+
+<img src="images/smartwilds-bioacoustics/deployment.jpg" alt="Song Meter Micro 2 recorder zip-tied to a branch, labeled SM03" width="520">
+
+*A Wildlife Acoustics Song Meter Micro 2 (SM03) deployed in a tree at The Wilds.*
+
+**Every retained acoustic recording from 10 recorder deployments at The Wilds in
+summer 2025, the schedule each recorder actually ran, and the birds BirdNET
+heard in each file.**
+
+This is the bioacoustics dataset of the **SmartWilds** collection. Each modality
+has its own dataset and DOI. They share one Darwin Core Event hierarchy rooted
+at `TheWilds_2025` and are joined by the
+[SmartWilds collection](https://huggingface.co/collections/ABC-Center/smartwilds-6af3c0aac0a67864cbf6edb4).
+
+## Dataset Details
+
+- **Curated by:** Jenna M. Kline, Anirudh Potlapally, Elizabeth G. Campolongo
+- **Homepage:** <https://github.com/Imageomics/smartwilds> (documentation: <https://imageomics.github.io/smartwilds/>)
+- **Paper:** SmartWilds: Integrating, Aligning, and Documenting Multi-Modal Field Sensing Data (under review)
+- **Release build:** 2026-10-22
+
+| Campaign | Dates | Deployments | WAV released |
+|---|---|---:|---:|
+| 1 | 2025-06-30 to 2025-07-03 | 4 | 311 |
+| 2 | 2025-08-04 to 2025-08-28 | 6 | 2,060 |
+
+**Darwin Core tables.** The Event table has 2,382 rows: 1 `Project`, 10
+`SensorDeployment`, and 2,371 `AudioRecording` Events, one per WAV file with its
+start time. There are 9,960 Occurrences. Each Occurrence is one
+**recording × species** pair, and `individualCount` gives the number of 3-second
+BirdNET detections of that species in the file (93,509 in total).
+
+### Supported Tasks
+
+- **Bird audio classification and re-labeling:** the raw WAV files plus the
+  BirdNET settings let you rerun BirdNET or swap in another classifier.
+- **Duty-cycle and acquisition-workload studies:** each recorder ran a different
+  schedule (5 min/h, 15 min/h, sunrise-sunset continuous, 1 h at dawn and
+  dusk), so `metadata/deployments.csv` supports counterfactual sampling-effort
+  analyses.
+- **Cross-modal comparison:** recorders sit beside the camera traps at
+  TW02, TW06, and TW09.
+
+## Dataset Structure
+
+```
+smartwilds-bioacoustics/
+├── data/
+│   ├── events.csv          # Project, SensorDeployment, AudioRecording Events
+│   ├── occurrences.csv     # one row per recording x species
+│   └── recordings.csv      # one row per WAV path on disk (2,453): disposition
+├── media/                  # media/<campaign>/<deployment>/<sd-card>/*.wav
+├── media_manifest.csv
+├── metadata/
+│   ├── deployments.csv     # recorder, site, GPS, schedule, WAV counts
+│   ├── birdnet_settings.csv
+│   └── provenance/         # raw-WAV reconciliation report, repair audit
+└── scripts/
+```
+
+`associatedMedia` on each `AudioRecording` Event and on each Occurrence gives
+the release-relative path of the WAV file.
+
+### Sensor settings
+
+All recorders are Wildlife Acoustics **Song Meter Micro 2**, 48 kHz.
+`metadata/deployments.csv` records the schedule each recorder actually ran.
+TW05-SM01 in Campaign 1 was configured for 1 h at sunrise and sunset but
+recorded 5-minute clips every hour. The release documents the actual schedule
+and notes the intended one.
+
+| Schedule | Duty cycle | Recorders |
+|---|---|---|
+| 5 min every hour | 8.3% | TW05-SM01, TW06-SM03, TW08-SM04 (C1); TW02-SM04 (C2) |
+| 15 min every hour | 25% | TW06-SM05 (C2) |
+| Sunrise to sunset, continuous | ~50-60% | TW05-SM06, TW07-SM02, TW08-SM03, TW09-SM01 (C2) |
+| 1 h at sunrise and sunset | ~8% | TW07-SM02 (C1) |
+
+### Annotation method
+
+- **Method:** automated. No human verification (all rows are `unverified`).
+- **Classifier:** BirdNET Analyzer v2.4, with 3.0 s segments, a 48 kHz sample
+  rate, and a 0-15 kHz bandpass.
+- **Taxon fields:** filled from BirdNET common-name mappings.
+- **Coordinates:** the recorder location, not the animal location.
+
+## Dataset Creation
+
+### Raw-WAV reconciliation
+
+There are 2,453 WAV paths on disk. `data/recordings.csv` classifies every one:
+
+| Disposition | Files |
+|---|---:|
+| Released, with detections | 2,307 |
+| Released, BirdNET ran but detected nothing (Event, no Occurrence) | 64 |
+| Excluded: Campaign-1 files copied into a Campaign-2 directory (transfer duplicates) | 78 |
+| Excluded: no BirdNET result, either truncated (3.9-22 s) or unreadable | 4 |
+
+The zero-detection recordings are kept as Events. This lets a user tell
+"recorded, nothing heard" apart from "not recorded".
+
+## Considerations for Using the Data
+
+- **`individualCount` counts detections, not birds.** A singing bird produces
+  many 3-second detections.
+- **No confidence threshold is recorded per Occurrence.** Re-run BirdNET from the
+  WAV files if you need score-level control.
+- **Acquisition is driven by the schedule.** Arrival burstiness is low
+  (B ≈ +0.10). Detection burstiness is high (B ≈ +0.53) and concentrated in the
+  dawn chorus.
+
+## Supplemental Material
+
+Extended inventories, the timestamp-repair methods, the standards crosswalk,
+validation checks, and the full defect inventory are in
+[`SUPPLEMENTAL.md`](https://huggingface.co/datasets/ABC-Center/smartwilds_metadata/blob/main/SUPPLEMENTAL.md)
+in the study-level SmartWilds dataset.
+
+## Licensing
+
+Data: [CC0 1.0](https://creativecommons.org/publicdomain/zero/1.0/). Scripts: MIT.
+
+## Citation
+
+```bibtex
+@misc{smartwilds2025_bioacoustics,
+  author    = {Kline, Jenna M. and Potlapally, Anirudh and Campolongo, Elizabeth G.},
+  title     = {SmartWilds Bioacoustics (The Wilds 2025)},
+  year      = {2026},
+  publisher = {Hugging Face},
+  note      = {DOI pending the raw data release},
+  url       = {https://huggingface.co/datasets/ABC-Center/smartwilds_metadata/tree/main/smartwilds-bioacoustics}
+}
+```
+
+Please also cite the paper:
+
+```bibtex
+@article{kline2026smartwilds,
+  author  = {Kline, Jenna M. and Potlapally, Anirudh and Campolongo, Elizabeth G.},
+  title   = {SmartWilds: Integrating, Aligning, and Documenting Multi-Modal Field Sensing Data},
+  year    = {2026},
+  note    = {Under review}
+}
+```
+
+This dataset card follows the [FAIR² Drones Data Standard](https://imageomics.github.io/fair_drones/)
+(Kline and Campolongo, 2026).
+
+## Acknowledgements
+
+This project is supported by the [Imageomics Institute](https://imageomics.org), which is funded by the US National Science Foundation's Harnessing the Data Revolution (HDR) program under [Award #2118240](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2118240) (Imageomics: A New Frontier of Biological Information Powered by Knowledge-Guided Machine Learning), and by the [AI Institute for Intelligent Cyberinfrastructure with Computational Learning in the Environment (ICICLE)](https://icicle.osu.edu/), which is funded by the US National Science Foundation under [Award #2112606](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2112606).
+
+This work is also supported by the AI and Biodiversity Change (ABC) Global Center. The ABC Global Center is funded by the US National Science Foundation under [Award No. 2330423](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2330423&HistoricalAwards=false) and the Natural Sciences and Engineering Research Council of Canada under [Award No. 585136](https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=782440). This dataset draws on research supported by the Social Sciences and Humanities Research Council.
+
+Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation, Natural Sciences and Engineering Research Council of Canada, or Social Sciences and Humanities Research Council.
+
+Special thanks to Dan Beetem and [The Wilds](https://www.thewilds.org/) in Ohio for providing access to deployment sites and supporting this wildlife monitoring research. We thank the Columbus Zoo and Aquarium for their support in facilitating this project. All data collection, including drone flights, was conducted under the supervision of the Director of Animal Management, with permission from The Wilds Animal Care and Use Committee.
+
+This work used the [Ohio Supercomputer Center](https://www.osc.edu/) (Ohio Supercomputer Center. 1987. Ohio Supercomputer Center. Columbus OH: Ohio Supercomputer Center. <https://ror.org/01apna436>).
+
+## Dataset Card Contact
+
+Jenna M. Kline (kline.377@osu.edu)

smartwilds-bioacoustics/metadata/DATA_DICTIONARY.md

@@ -0,0 +1,147 @@
+# smartwilds-bioacoustics data dictionary
+
+Generated by `scripts/build_hf_release.py`. Darwin Core term definitions: <https://dwc.tdwg.org/terms/>. Columns not listed in a Darwin Core namespace are SmartWilds-specific.
+
+## `data/events.csv`
+
+2,382 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `eventID` | Darwin Core: unique Event identifier | TheWilds_2025_bioacoustics |
+| `parentEventID` | Darwin Core: eventID of the parent Event | TheWilds_2025 |
+| `eventType` | Darwin Core: Project, SensorDeployment, AudioRecording, MediaEvent, or ObservationSession | Project |
+| `eventDate` | Darwin Core: ISO 8601 date or date interval (start/end) | 2025-06-30/2025-08-28 |
+| `samplingProtocol` | Darwin Core: sensor, schedule, and processing chain | Passive acoustic monitoring using Wildlife Acoustics Song Me… |
+| `sampleSizeValue` | Darwin Core: sample size | 1 |
+| `sampleSizeUnit` | Darwin Core: unit of sampleSizeValue | WAV recording |
+| `locality` | Darwin Core: named place | The Wilds Conservation Center, Muskingum County, Ohio, USA |
+| `locationID` | Darwin Core: site (TWxx) for fixed sensors; flight ID for drones | TW05 |
+| `decimalLatitude` | Darwin Core: WGS84; see dataGeneralizations for what the point represents | 39.83 |
+| `decimalLongitude` | Darwin Core: WGS84 | -81.74 |
+| `coordinateUncertaintyInMeters` | Darwin Core: meters | 5000 |
+| `geodeticDatum` | Darwin Core: WGS84 | WGS84 |
+| `country` |  | United States |
+| `stateProvince` |  | Ohio |
+| `county` |  | Muskingum |
+| `datasetID` | Darwin Core: modality dataset identifier | TheWilds_2025_bioacoustics |
+| `datasetName` | Darwin Core | The Wilds 2025 Multi-Modal Survey - Bioacoustics |
+| `recordedBy` | Darwin Core: device or team | Tanishka Wani; Jenna Kline |
+| `fieldNotes` | Darwin Core: provenance and field notes | 10 Song Meter Micro 2 deployments across 6 sites, 2 campaign… |
+| `eventTime` | Darwin Core: local time (or interval); blank when not recoverable | 07:48:02 |
+| `associatedMedia` | Darwin Core: release-relative path to the source media file | media/thewilds_20250804_20250828/TW09-SM01/Data/SM01_2025082… |
+
+## `data/occurrences.csv`
+
+9,960 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `occurrenceID` | Darwin Core: unique Occurrence identifier | TheWilds_2025_ARU_000001 |
+| `eventID` | Darwin Core: unique Event identifier | TheWilds_2025_ARU_SM01_20250824_074802 |
+| `catalogNumber` | Darwin Core: source record identifier | ARU_SM01_20250824_074802_Belted_Kingfisher |
+| `scientificName` | Darwin Core: top-1 model or annotator taxon | Megaceryle alcyon |
+| `kingdom` | Darwin Core: rank | Animalia |
+| `phylum` | Darwin Core: rank | Chordata |
+| `class` | Darwin Core: rank | Aves |
+| `order` | Darwin Core: rank |  |
+| `family` | Darwin Core: rank |  |
+| `genus` | Darwin Core: rank | Megaceryle |
+| `specificEpithet` | Darwin Core: rank | alcyon |
+| `taxonRank` | Darwin Core: rank of scientificName | species |
+| `vernacularName` | Darwin Core: common name | Belted Kingfisher |
+| `occurrenceStatus` | Darwin Core: present | present |
+| `individualCount` | Darwin Core: 1 per visual crop; number of BirdNET detections for acoustic rows | 662 |
+| `basisOfRecord` | Darwin Core: MachineObservation or HumanObservation | MachineObservation |
+| `recordedBy` | Darwin Core: device or team | Song Meter Micro 2 (SM01) |
+| `identifiedBy` | Darwin Core: model (and version) or annotator | BirdNET Analyzer v2.4 |
+| `identificationVerificationStatus` | Darwin Core: verification status; the vocabulary differs by table (SUPPLEMENTAL.md S9.3) | unverified |
+| `identificationRemarks` | Darwin Core: rank, top-1 score, bbox xywh (px), detector label | Automated classification by BirdNET; 662 detection(s) in rec… |
+| `eventDate` | Darwin Core: ISO 8601 date or date interval (start/end) | 2025-08-24 |
+| `eventTime` | Darwin Core: local time (or interval); blank when not recoverable | 07:48:02 |
+| `decimalLatitude` | Darwin Core: WGS84; see dataGeneralizations for what the point represents | 39.83418 |
+| `decimalLongitude` | Darwin Core: WGS84 | -81.73783 |
+| `coordinateUncertaintyInMeters` | Darwin Core: meters | 50 |
+| `geodeticDatum` | Darwin Core: WGS84 | WGS84 |
+| `locality` | Darwin Core: named place | The Wilds - Dock |
+| `locationID` | Darwin Core: site (TWxx) for fixed sensors; flight ID for drones | TW09 |
+| `country` |  | United States |
+| `stateProvince` |  | Ohio |
+| `county` |  | Muskingum |
+| `establishmentMeans` |  | wild |
+| `degreeOfEstablishment` |  | native |
+| `occurrenceRemarks` | Darwin Core: projection status, timestamp method, repair notes (key=value; ...) |  |
+| `associatedMedia` | Darwin Core: release-relative path to the source media file | media/thewilds_20250804_20250828/TW09-SM01/Data/SM01_2025082… |
+| `datasetID` | Darwin Core: modality dataset identifier | TheWilds_2025_bioacoustics |
+| `datasetName` | Darwin Core | The Wilds 2025 Multi-Modal Survey - Bioacoustics |
+
+## `data/recordings.csv`
+
+2,453 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `release_path` | Path of the file inside this dataset | media/thewilds_20250630_20250703/TW05-SM01/SD01_20250630_202… |
+| `raw_wav_path` |  | thewilds_20250630_20250703/bioacoustics/TW05-SM01/SD01_20250… |
+| `source_campaign_directory` |  | campaign1 |
+| `deploymentID` |  | TW05-SM01 |
+| `recording_file` |  | SM001_20250630_195900.wav |
+| `recording_date` |  | 2025-06-30 |
+| `recording_time` |  | 19:59:00 |
+| `in_source_campaign_window` |  | true |
+| `wav_bytes` |  | 14401138 |
+| `wav_header_status` |  | not_checked_release_represented_or_transfer |
+| `duration_seconds` |  | 12.629333 |
+| `birdnet_result_path` |  | BioacousticAnalysis/BirdnetCSV_SunriseSunset/SM001_20250630_… |
+| `birdnet_result_status` |  | nonempty |
+| `darwin_core_eventID` |  | TheWilds_2025_ARU_SM01_20250630_195900 |
+| `reconciliation_status` |  | represented_recording |
+| `release_disposition` |  | included_as_AudioRecording_Event |
+
+## `media_manifest.csv`
+
+2,475 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `release_path` | Path of the file inside this dataset | media/thewilds_20250630_20250703/TW05-SM01/SD01_20250630_202… |
+| `source_path` | Path in the project archive | thewilds_20250630_20250703/bioacoustics/TW05-SM01/SD01_20250… |
+| `bytes` | File size | 14401138 |
+| `include` | true if the file is distributed in media/ | true |
+| `exclusion_reason` | Why a file is not distributed | excluded_from_release:transfer_duplicate |
+
+## `metadata/birdnet_settings.csv`
+
+1 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `software` |  | BirdNET Analyzer |
+| `version` |  | 2.4 |
+| `segment_length_s` |  | 3.0 |
+| `sample_rate_hz` |  | 48000 |
+| `bandpass_hz` |  | 0-15000 |
+| `occurrence_unit` |  | one row per recording x species; individualCount = detection… |
+
+## `metadata/deployments.csv`
+
+10 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `deploymentID` |  | TW05-SM01 |
+| `model` |  | Wildlife Acoustics Song Meter Micro 2 |
+| `Recorder` |  | SM01 |
+| `Site` |  | TW05 - Lake Trail Gate O |
+| `GPS` |  | 39.83537, -81.73842 |
+| `Orientation` |  | East |
+| `Height` |  | ~1.5m |
+| `Recording Schedule` |  | 5-min clips every hour (intended 1hr sunrise/sunset — config… |
+| `WAV Count` |  | 144 |
+| `Storage` |  | — |
+| `campaign` |  | thewilds_20250630_20250703 |
+| `Deploy` |  | Aug 4 |
+| `Retrieve` |  | Aug 28 |
+| `wav_paths` |  | 144 |
+| `wav_released` |  | 144 |
+

smartwilds-bioacoustics/metadata/birdnet_settings.csv

@@ -0,0 +1,2 @@
+software,version,segment_length_s,sample_rate_hz,bandpass_hz,occurrence_unit
+BirdNET Analyzer,2.4,3.0,48000,0-15000,one row per recording x species; individualCount = detections

smartwilds-bioacoustics/metadata/deployments.csv

@@ -0,0 +1,11 @@
+deploymentID,model,Recorder,Site,GPS,Orientation,Height,Recording Schedule,WAV Count,Storage,campaign,Deploy,Retrieve,wav_paths,wav_released
+TW05-SM01,Wildlife Acoustics Song Meter Micro 2,SM01,TW05 - Lake Trail Gate O,"39.83537, -81.73842",East,~1.5m,5-min clips every hour (intended 1hr sunrise/sunset — config error),144,—,thewilds_20250630_20250703,,,144,144
+TW06-SM03,Wildlife Acoustics Song Meter Micro 2,SM03,TW06 - Zebra Shelter,"39.83518, -81.73087",Southeast,2m,"5 min every hour, top of hour, 24/7",77,—,thewilds_20250630_20250703,,,77,77
+TW07-SM02,Wildlife Acoustics Song Meter Micro 2,SM02,TW07 - Zipline Tower,"39.83288, -81.73455",Southwest,~1.5m,1hr at sunrise + 1hr at sunset,12,—,thewilds_20250630_20250703,,,12,12
+TW08-SM04,Wildlife Acoustics Song Meter Micro 2,SM04,TW08 - Fence w/ Vegetation,"39.82900, -81.74378",Southwest,~1.5m,"5 min every hour, top of hour, 24/7",78,—,thewilds_20250630_20250703,,,78,78
+TW02-SM04,Wildlife Acoustics Song Meter Micro 2,SM04,TW02 - Nomad Ridge Shelter,"39.82957, -81.73295",,,5 min every hour,659,26.06 GB,thewilds_20250804_20250828,Aug 4,Aug 28,659,581
+TW05-SM06,Wildlife Acoustics Song Meter Micro 2,SM06,TW05 - Lake Trail Gate O,"39.82866, -81.73758",,,Sunrise to sunset (continuous),100,16.62 GB,thewilds_20250804_20250828,Aug 4,Aug 28,100,96
+TW06-SM05,Wildlife Acoustics Song Meter Micro 2,SM05,TW06 - Zebra Shelter,"39.82870, -81.73756",,,15 min every hour,508,21.82 GB,thewilds_20250804_20250828,Aug 4,Aug 28,508,508
+TW07-SM02,Wildlife Acoustics Song Meter Micro 2,SM02,TW07 - Zipline Tower,"39.82891, -81.74392",,,Sunrise to sunset (continuous),339,56.44 GB,thewilds_20250804_20250828,Aug 4,Aug 28,339,339
+TW08-SM03,Wildlife Acoustics Song Meter Micro 2,SM03,TW08 - Fence w/ Vegetation,"39.82892, -81.74392",,,Sunrise to sunset (continuous),341,56.69 GB,thewilds_20250804_20250828,Aug 4,Aug 28,341,341
+TW09-SM01,Wildlife Acoustics Song Meter Micro 2,SM01,TW09 - Dock,"39.83418, -81.73783",,,Sunrise to sunset (continuous),195,32.38 GB,thewilds_20250804_20250828,Aug 14,Aug 28,195,195

smartwilds-bioacoustics/metadata/provenance/RAW_WAV_RECONCILIATION_20261008.md

@@ -0,0 +1,40 @@
+# Raw WAV Reconciliation
+
+## Outcome
+
+The corrected release has **2,371 unique `AudioRecording` Events**. It contains
+no duplicate Event for transferred raw media and keeps the existing 9,960
+BirdNET Occurrences unchanged.
+
+| Raw-file disposition | Files | Release treatment |
+| --- | ---: | --- |
+| Represented recording | 2307 | One `AudioRecording` Event |
+| Processed, zero detections | 64 | One `AudioRecording` Event; no Occurrence |
+| Transfer duplicate | 78 | Excluded; source belongs to another campaign/deployment |
+| Truncated/corrupt, unprocessed | 4 | Excluded; no BirdNET result |
+| **Raw files on disk** | **2453** | **Reconciled** |
+
+## Why The Earlier Count Was Misleading
+
+The former comparison of 2,453 raw WAV files with 2,307 Events treated every
+physical file path as a separate capture. Seventy-eight Campaign-1 WAVs were
+transferred into the Campaign-2 `TW02-SM04/Data` directory; they are duplicate
+transfers, not new Campaign-2 recordings. The remaining 68 unmatched
+in-window files divide into 64 successfully processed files with header-only
+BirdNET results (zero detections) and four truncated/corrupt files with no
+BirdNET result.
+
+The 64 zero-detection recordings are now represented as Events with a field
+note identifying the header-only BirdNET result. They deliberately generate no
+Occurrence. The four failures remain out of the Event table and are listed in
+the accompanying CSV with their header status and duration.
+
+## Integrity Checks
+
+- Every `AudioRecording` Event has a unique recording key: deployment, recorder,
+  date, and time.
+- The 78 transfer duplicates never create a second Event.
+- All 9,960 Occurrences retain valid Event links.
+- The CSV provides one row per raw WAV path, including path, deployment, file size, BirdNET-result status, final Event ID,
+  and release disposition. Header status and duration are recorded for the four
+  unrepresented files that require failure evidence.

smartwilds-bioacoustics/metadata/provenance/bioacoustic_event_repair_20261008.json

@@ -0,0 +1,18 @@
+{
+  "repair": "bioacoustic AudioRecording completeness and raw-WAV reconciliation",
+  "snapshot": "integration_pass_20261022",
+  "raw_wav_files": 2453,
+  "audio_recording_events": 2371,
+  "occurrences_unchanged": 9960,
+  "reconciliation_status_counts": {
+    "processed_zero_detection": 64,
+    "represented_recording": 2307,
+    "transfer_duplicate": 78,
+    "truncated_or_corrupt_unprocessed": 4
+  },
+  "release_policy": {
+    "transfer_duplicate": "excluded; no duplicate Event",
+    "processed_zero_detection": "included as Event with no Occurrence",
+    "truncated_or_corrupt_unprocessed": "excluded; documented in reconciliation CSV"
+  }
+}

smartwilds-bioacoustics/scripts/build_hf_release.py

@@ -0,0 +1,716 @@
+#!/usr/bin/env python3
+"""Stage the SmartWilds per-modality Hugging Face dataset directories.
+
+One directory per modality (camera traps, bioacoustics, drones, citizen
+science) plus the umbrella ``smartwilds`` dataset, each laid out like
+imageomics/kabr-behavior-telemetry:
+
+    <repo>/README.md                 dataset card (filled in separately)
+    <repo>/data/events.csv           Darwin Core Event table
+    <repo>/data/occurrences.csv      Darwin Core Occurrence table
+    <repo>/data/...                  modality-specific per-file tables
+    <repo>/metadata/                 sensor settings, data dictionary, provenance
+    <repo>/scripts/                  generator / georeferencing / this script
+    <repo>/media_manifest.csv        one row per raw file: release path, bytes,
+                                     include flag, exclusion reason
+
+Inputs are the frozen Darwin Core tables of the integration pass
+(``SMARTWILDS_DWC_SNAPSHOT``) and the validated post-triage camera derivative
+(``SMARTWILDS_CAMERA_SOURCE``). Neither is modified. The staged copies get one
+release repair, audited in ``RELEASE_BUILD_AUDIT.json``: ``associatedMedia`` is
+rewritten from absolute archive paths to release-relative ``media/...`` paths,
+and drone frames in the purged scratch workspace fall back to their retained
+original video.
+
+Media files are NOT copied here. ``media_manifest.csv`` is the packaging input.
+
+Locations are read from environment variables, all optional:
+
+    SMARTWILDS_ARCHIVE         field archive root (default: this repository)
+    SMARTWILDS_DWC_SNAPSHOT    frozen Darwin Core tables
+    SMARTWILDS_CAMERA_SOURCE   validated camera derivative
+    SMARTWILDS_RELEASE         output directory (also ``--out``)
+    SMARTWILDS_PUBLIC_SCRIPTS  the public scripts shipped inside each dataset
+    SMARTWILDS_SCRATCH_PREFIX  prefix of purged scratch frame paths in the tables
+    SMARTWILDS_DETECTIONS      prefix of the external detection workspace
+
+Usage:
+    python release/scripts/build_hf_release.py [--out DIR] [--only REPO ...]
+"""
+import argparse
+import hashlib
+import json
+import os
+import re
+import shutil
+import subprocess
+import sys
+from datetime import datetime, timezone
+from pathlib import Path
+
+import pandas as pd
+
+def env_path(name, default):
+    return Path(os.environ.get(name, default)).resolve()
+
+
+ROOT = env_path("SMARTWILDS_ARCHIVE", Path(__file__).resolve().parents[2])
+SNAP = env_path("SMARTWILDS_DWC_SNAPSHOT", ROOT / "processing_output/darwin_core_latest_20261008")
+CAMERA_SOURCE = env_path("SMARTWILDS_CAMERA_SOURCE", ROOT / "archive/camera_traps_validated_20261108")
+RELEASE_DIR = env_path("SMARTWILDS_RELEASE", ROOT / "release/hf_smartwilds_20261022")
+PUBLIC_SCRIPTS = env_path("SMARTWILDS_PUBLIC_SCRIPTS", ROOT / "public/smartwilds/scripts")
+TS_MAP = ROOT / "processing_output/detections/drone_camera_analysis/camera_time_mapping"
+SENSOR_MD = ROOT / "documentation/sensor_configurations.md"
+CARDS = ROOT / "release/cards"
+CAMPAIGNS = ["thewilds_20250630_20250703", "thewilds_20250804_20250828", "thewilds_20250908_20250908"]
+PROJECT_PREFIX = str(ROOT) + "/"
+# Prefixes of absolute paths that may still appear inside the frozen tables.
+SCRATCH_PREFIX = os.environ.get("SMARTWILDS_SCRATCH_PREFIX", "")      # purged drone frame workspace
+DETECTIONS_PREFIX = os.environ.get("SMARTWILDS_DETECTIONS", "")        # external detection workspaces
+if DETECTIONS_PREFIX and not DETECTIONS_PREFIX.endswith("/"):
+    DETECTIONS_PREFIX += "/"
+# Anything under the same filesystem roots as the known prefixes counts as an
+# archive path (e.g. "/fs/ess/" for "/fs/ess/<project>/TheWilds2025/").
+CLUSTER_MARKERS = sorted({"/".join(m.split("/")[:3]) + "/" for m in (PROJECT_PREFIX, SCRATCH_PREFIX, DETECTIONS_PREFIX)
+                          if m and m.startswith("/")})
+STUDY_START, STUDY_END = "2025-06-30", "2025-09-08"
+
+REPOS = {
+    "camera_traps": "smartwilds-camera-traps",
+    "bioacoustics": "smartwilds-bioacoustics",
+    "drones": "smartwilds-drones",
+    "citizen_science": "smartwilds-citizen-science",
+    "core": "smartwilds",
+}
+
+audit = {"inputs": {"snapshot": str(SNAP), "camera_source": str(CAMERA_SOURCE)}, "built_utc": None,
+         "repos": {}, "blocked": [], "errors": []}
+
+
+def read(path, **kw):
+    return pd.read_csv(path, dtype=str, keep_default_na=False, **kw)
+
+
+def readable(path):
+    return os.access(path, os.R_OK)
+
+
+def sha256(path, buf=1 << 20):
+    h = hashlib.sha256()
+    with open(path, "rb") as f:
+        for chunk in iter(lambda: f.read(buf), b""):
+            h.update(chunk)
+    return h.hexdigest()
+
+
+def rel_media(abs_path):
+    """<archive>/<campaign>/<modality>/<rest> -> media/<campaign>/<rest>."""
+    if not abs_path.startswith(PROJECT_PREFIX):
+        return None
+    parts = abs_path[len(PROJECT_PREFIX):].split("/")
+    if parts[0] not in CAMPAIGNS or len(parts) < 3:
+        return None
+    return "/".join(["media", parts[0]] + parts[2:])
+
+
+BIRD_WATER = ROOT / "processing_output/detections/drone_camera_analysis/aquatic_bird_analysis/bird_water_association_per_box.csv"
+PATH_REWRITES = [(str(BIRD_WATER), "analysis/bird_water_association_per_box.csv")]
+if DETECTIONS_PREFIX:
+    PATH_REWRITES.append((DETECTIONS_PREFIX, "external_detections/"))
+PATH_REWRITES.append((PROJECT_PREFIX, "TheWilds2025/"))
+
+
+def has_cluster_path(col):
+    return any(col.str.contains(m, regex=False).any() for m in CLUSTER_MARKERS)
+
+
+def count_cluster_paths(df):
+    return int(sum(df[c].str.contains(m, regex=False).sum() for c in df.columns for m in CLUSTER_MARKERS))
+
+
+def strip_cluster_paths(df):
+    """Rewrite remaining absolute archive paths in free-text columns to
+    release-relative (shipped) or project-relative (provenance-only) forms."""
+    for c in df.columns:
+        col = df[c]
+        if has_cluster_path(col):
+            for a, b in PATH_REWRITES:
+                col = col.str.replace(a, b, regex=False)
+            df[c] = col
+    return df
+
+
+def scrub_text_copy(src, dst, extra=()):
+    """Copy a text file (JSON, Markdown), rewriting archive paths and the
+    provenance tokens of the camera derivative."""
+    text = Path(src).read_text()
+    rewrites = list(PATH_REWRITES) + list(ARRIVAL_TOKEN_REWRITES) + [(SNAP.name, "integration_pass_20261022")] + list(extra)
+    for a, b in rewrites:
+        text = text.replace(a, b)
+    Path(dst).write_text(text)
+
+
+# Files written by earlier builds that the current build no longer produces.
+STALE = ["scripts/generate_darwin_core_latest_20261008.py", "scripts/refresh_icde_camera_arrival_metadata.py",
+         "analysis/arrival_traces_v2/DEPRECATED_LEGACY_PRIMITIVES.md",
+         "analysis/arrival_traces_v2/UNAVAILABLE_CAMERA_IMAGE_CAPTURE_ACQUISITION.md",
+         "analysis/bird_water_association_per_box.csv"]
+
+
+def atomic_write(path, writer):
+    """Write through a temp file and rename, so files owned by another team
+    member (and read-only to us) are replaced rather than opened for writing."""
+    path.parent.mkdir(parents=True, exist_ok=True)
+    temp = path.with_name(f".{path.name}.tmp")
+    try:
+        writer(temp)
+        temp.replace(path)
+    finally:
+        temp.unlink(missing_ok=True)
+
+
+def parse_md_tables(md_path):
+    """Return {"## section / ### heading": DataFrame} for every pipe table."""
+    tables, section, heading, rows = {}, "", None, []
+    for line in Path(md_path).read_text().splitlines() + [""]:
+        if line.startswith("## "):
+            section = heading = line[3:].strip()
+        elif line.startswith("### "):
+            heading = section + " / " + line[4:].strip()
+        if line.startswith("|"):
+            cells = [c.strip() for c in line.strip().strip("|").split("|")]
+            if not all(re.fullmatch(r":?-+:?", c) for c in cells):
+                rows.append(cells)
+        elif rows:
+            tables[heading] = pd.DataFrame(rows[1:], columns=rows[0])
+            rows = []
+    return tables
+
+
+def media_inventory(modality, exclude=lambda rel: None):
+    """One row per raw file under every campaign's <modality>/ directory."""
+    rows = []
+    for camp in CAMPAIGNS:
+        base = ROOT / camp / modality
+        if not base.is_dir():
+            continue
+        for dirpath, _, files in os.walk(base):
+            for fn in files:
+                src = Path(dirpath) / fn
+                rel = rel_media(str(src))
+                reason = exclude(rel) or ""
+                try:
+                    size = src.stat().st_size
+                except OSError:
+                    size, reason = "", reason or "unreadable_source"
+                rows.append({"release_path": rel, "source_path": str(src.relative_to(ROOT)),
+                             "bytes": size, "include": "false" if reason else "true",
+                             "exclusion_reason": reason})
+    return pd.DataFrame(rows).sort_values("release_path").reset_index(drop=True)
+
+
+def check_tables(name, ev, oc, manifest, extra_event_ids=()):
+    """Release checks the snapshot audit lacked: dates inside parent spans,
+    sentinel strings, and every associatedMedia resolving to an included file."""
+    res = {}
+    ids = set(ev.eventID) | set(extra_event_ids)
+    res["events"], res["occurrences"] = len(ev), len(oc)
+    res["dangling_parentEventID"] = sorted({p for p in ev.parentEventID if p and p not in ids})
+    res["occurrences_without_event"] = int((~oc.eventID.isin(ids)).sum())
+    sentinel = re.compile(r"^(nan|none|null|-1|1970-01-01)$", re.I)
+    res["sentinel_cells"] = {f"{t}.{c}": int(df[c].str.match(sentinel).sum())
+                             for t, df in (("events", ev), ("occurrences", oc)) for c in df.columns
+                             if df[c].str.match(sentinel).any()}
+    span = ev.set_index("eventID").eventDate.str.split("/", expand=True)
+    span.columns = ["a", "b"][: span.shape[1]]
+    if "b" not in span:
+        span["b"] = None
+    span["b"] = span.b.fillna(span.a)
+    single = oc[~oc.eventDate.str.contains("/")]
+    j = single.join(span, on="eventID")
+    res["occurrence_date_outside_event_span"] = int(((j.eventDate < j.a) | (j.eventDate > j.b)).sum())
+    res["occurrence_date_outside_study"] = int(((single.eventDate < STUDY_START) | (single.eventDate > STUDY_END)).sum())
+    res["occurrence_interval_dates"] = int(oc.eventDate.str.contains("/").sum())
+    media = oc.associatedMedia[oc.associatedMedia != ""]
+    res["absolute_media_paths"] = int(media.str.startswith("/").sum())
+    res["cells_with_cluster_paths"] = count_cluster_paths(ev) + count_cluster_paths(oc)
+    if manifest is not None:
+        included = set(manifest.release_path[manifest.include == "true"])
+        unresolved = media[~media.isin(included)]
+        res["media_paths_not_in_included_manifest"] = int(len(unresolved))
+        res["media_unresolved_examples"] = unresolved.drop_duplicates().head(5).tolist()
+    res["pass"] = (not res["dangling_parentEventID"] and res["occurrences_without_event"] == 0
+                   and not res["sentinel_cells"] and res["occurrence_date_outside_event_span"] == 0
+                   and res["occurrence_date_outside_study"] == 0 and res["absolute_media_paths"] == 0
+                   and res["cells_with_cluster_paths"] == 0
+                   and res.get("media_paths_not_in_included_manifest", 0) == 0)
+    return res
+
+
+def write_repo(out, repo, files):
+    d = out / repo
+    for rel, obj in files.items():
+        p = d / rel
+        p.parent.mkdir(parents=True, exist_ok=True)
+        if isinstance(obj, pd.DataFrame):
+            atomic_write(p, lambda t: obj.to_csv(t, index=False))
+        elif isinstance(obj, (dict, list)):
+            atomic_write(p, lambda t: t.write_text(json.dumps(obj, indent=2) + "\n"))
+        elif isinstance(obj, Path) and obj.suffix in (".json", ".md"):
+            atomic_write(p, lambda t: scrub_text_copy(obj, t))   # provenance records: no archive paths or snapshot names
+        elif isinstance(obj, Path):
+            atomic_write(p, lambda t: shutil.copyfile(obj, t))
+        else:
+            atomic_write(p, lambda t: t.write_text(obj))
+    return d
+
+
+TERMS = {
+    # Darwin Core / Humboldt terms (https://dwc.tdwg.org/terms/)
+    "eventID": "Darwin Core: unique Event identifier", "parentEventID": "Darwin Core: eventID of the parent Event",
+    "eventType": "Darwin Core: Project, SensorDeployment, AudioRecording, MediaEvent, or ObservationSession",
+    "eventDate": "Darwin Core: ISO 8601 date or date interval (start/end)",
+    "eventTime": "Darwin Core: local time (or interval); blank when not recoverable",
+    "samplingProtocol": "Darwin Core: sensor, schedule, and processing chain", "sampleSizeValue": "Darwin Core: sample size",
+    "sampleSizeUnit": "Darwin Core: unit of sampleSizeValue", "fieldNotes": "Darwin Core: provenance and field notes",
+    "occurrenceID": "Darwin Core: unique Occurrence identifier", "catalogNumber": "Darwin Core: source record identifier",
+    "scientificName": "Darwin Core: top-1 model or annotator taxon", "taxonRank": "Darwin Core: rank of scientificName",
+    "vernacularName": "Darwin Core: common name", "kingdom": "Darwin Core: rank", "phylum": "Darwin Core: rank",
+    "class": "Darwin Core: rank", "order": "Darwin Core: rank", "family": "Darwin Core: rank", "genus": "Darwin Core: rank",
+    "specificEpithet": "Darwin Core: rank", "occurrenceStatus": "Darwin Core: present",
+    "individualCount": "Darwin Core: 1 per visual crop; number of BirdNET detections for acoustic rows",
+    "basisOfRecord": "Darwin Core: MachineObservation or HumanObservation",
+    "recordedBy": "Darwin Core: device or team", "identifiedBy": "Darwin Core: model (and version) or annotator",
+    "identificationVerificationStatus": "Darwin Core: verification status; the vocabulary differs by table (SUPPLEMENTAL.md S9.3)",
+    "identificationRemarks": "Darwin Core: rank, top-1 score, bbox xywh (px), detector label",
+    "decimalLatitude": "Darwin Core: WGS84; see dataGeneralizations for what the point represents",
+    "decimalLongitude": "Darwin Core: WGS84", "coordinateUncertaintyInMeters": "Darwin Core: meters",
+    "geodeticDatum": "Darwin Core: WGS84", "locality": "Darwin Core: named place",
+    "locationID": "Darwin Core: site (TWxx) for fixed sensors; flight ID for drones",
+    "occurrenceRemarks": "Darwin Core: projection status, timestamp method, repair notes (key=value; ...)",
+    "associatedMedia": "Darwin Core: release-relative path to the source media file",
+    "associatedReferences": "Darwin Core: release-relative supporting analysis file",
+    "dataGeneralizations": "Darwin Core: what the coordinate represents", "informationWithheld": "Darwin Core",
+    "datasetID": "Darwin Core: modality dataset identifier", "datasetName": "Darwin Core",
+    "release_path": "Path of the file inside this dataset", "source_path": "Path in the project archive",
+    "bytes": "File size", "include": "true if the file is distributed in media/",
+    "exclusion_reason": "Why a file is not distributed",
+    "has_human_detection": "true if any Occurrence on this image is labelled Homo sapiens",
+    "occurrence_count": "Occurrences derived from this image",
+    "video_path": "Release-relative path of the raw video", "detection_id": "<video>_<frame>_<k> from the SAM3 run",
+    "frame_index": "0-based native frame index in the video (SAM3 sampled every 15th frame)",
+    "bbox_x": "Box left, pixels, in the native frame", "bbox_y": "Box top, pixels", "bbox_w": "Box width, pixels",
+    "bbox_h": "Box height, pixels", "top1_score": "BioCLIP Tree-of-Life top-1 score",
+    "georef_status": "Ground-projection status (ok | invalid_altitude | missing_drone_gps | horizon_ray | out_of_range)",
+}
+
+
+def write_dictionary(d, repo):
+    out = [f"# {repo} data dictionary", "",
+           "Generated by `scripts/build_hf_release.py`. Darwin Core term definitions: "
+           "<https://dwc.tdwg.org/terms/>. Columns not listed in a Darwin Core namespace are "
+           "SmartWilds-specific.", ""]
+    for p in sorted(d.rglob("*.csv")):
+        rel = p.relative_to(d)
+        df = pd.read_csv(p, dtype=str, keep_default_na=False, nrows=1000)
+        n = sum(1 for _ in open(p)) - 1
+        out += [f"## `{rel}`", "", f"{n:,} rows.", "", "| Column | Description | Example |", "|---|---|---|"]
+        for c in df.columns:
+            ex = next((v for v in df[c] if v), "")
+            ex = (ex[:60] + "…") if len(ex) > 60 else ex
+            out.append(f"| `{c}` | {TERMS.get(c, '')} | {ex.replace('|', '/')} |")
+        out.append("")
+    atomic_write(d / "metadata/DATA_DICTIONARY.md", lambda t: t.write_text("\n".join(out) + "\n"))
+
+
+def write_checksums(d):
+    lines = []
+    for p in sorted(d.rglob("*")):
+        if p.is_file() and p.name not in ("CHECKSUMS.sha256", "README.md") and "images" not in p.parts:
+            lines.append(f"{sha256(p)}  {p.relative_to(d)}")
+    atomic_write(d / "CHECKSUMS.sha256", lambda t: t.write_text("\n".join(lines) + "\n"))
+
+
+# --------------------------------------------------------------------------
+# Camera traps
+# --------------------------------------------------------------------------
+def scrub_csv_copy(src, dst):
+    """Copy a CSV, normalising provenance tokens of the camera derivative."""
+    text = Path(src).read_text()
+    for a, b in ARRIVAL_TOKEN_REWRITES:
+        text = text.replace(a, b)
+    Path(dst).write_text(text)
+
+
+ARRIVAL_TOKEN_REWRITES = [
+    ("_arrival_icde_camera_", "_arrival_camera_"),
+    ("one retained ICDE SAM3-geometry camera Occurrence", "one retained original-SAM3-geometry camera Occurrence"),
+    ("icde_sam3_geometry_only_derivative_20261108", "camera_release_20261108"),
+]
+
+
+def camera_deployments(tables, images):
+    """Per-deployment table: sensor configuration joined to the geolocation
+    config and the number of processed images on disk."""
+    c1 = tables["Camera Trap Configurations / Campaign 1: June 30 – July 3, 2025"].assign(campaign="thewilds_20250630_20250703")
+    c2 = tables["Camera Trap Configurations / Campaign 2: August 4 – August 28, 2025"].assign(campaign="thewilds_20250804_20250828")
+    dep = pd.concat([t for t in (c1, c2) if "Camera" in t.columns], ignore_index=True)
+    dep.insert(0, "deploymentID", dep.Site.str.extract(r"^(TW\d+)")[0] + "-" + dep.Camera)
+    geo = read(SNAP / "scripts/georeferencing/camera_geolocation_config.csv")
+    geo["camera_id"] = geo.camera_id.str.replace("_", "-")  # raw dir is TW09_CT02
+    dep = dep.merge(geo[["collection", "camera_id", "path_contains", "camera_lat", "camera_lon", "hfov_deg", "bearing_deg",
+                         "image_rotation_deg", "target_height_m", "config_source", "notes"]].rename(
+                        columns={"notes": "geolocation_notes"}),
+                    left_on=["campaign", "deploymentID"], right_on=["collection", "camera_id"], how="left"
+                    ).drop(columns=["collection", "camera_id"]).fillna({"path_contains": ""})
+    dep["images_on_disk"] = dep.apply(lambda r: int(((images.dataset == r.campaign) &
+                                                      (images.camera_id.str.replace("_", "-") == r.deploymentID)
+                                                      & (images.release_path.str.contains(r.path_contains, regex=False)
+                                                         if r.path_contains else True)).sum()), axis=1)
+    # A path-specific override (TW09-CT02 SD23, after the Aug 18 move) takes its
+    # images away from the deployment's default row.
+    for (camp, did), g in dep.groupby(["campaign", "deploymentID"]):
+        if len(g) > 1:
+            default = g.index[g.path_contains == ""]
+            dep.loc[default, "images_on_disk"] -= g.loc[g.path_contains != "", "images_on_disk"].sum()
+    dep["released"] = (dep.deploymentID != "TW07-CT05").map({True: "true", False: "false"})
+    return dep
+
+
+def build_camera(out, tables):
+    """Camera tables come from the validated post-triage derivative: the final
+    Event, Occurrence, image-inventory, media-manifest, taxonomy,
+    annotation-provenance, and arrival-analysis products."""
+    src = CAMERA_SOURCE / "smartwilds-camera-traps"
+    ev = read(src / "data/events.csv")
+    oc = read(src / "data/occurrences.csv")
+    images = read(src / "data/images.csv")
+    if len(ev) != 19785 or len(oc) != 72775 or len(images) != 235778:
+        raise RuntimeError("validated camera source count mismatch")
+    if set(oc.geometryProvenance) != {"SAM3"}:
+        raise RuntimeError("validated camera source contains non-SAM3 geometry")
+    if "behavior" in oc.columns or oc.astype(str).apply(
+            lambda col: col.str.contains(r"water|shore|aquatic", case=False, regex=True).any()).any():
+        raise RuntimeError("validated camera source retains water-association data")
+    if oc.associatedMedia.str.startswith("/").any():
+        raise RuntimeError("validated camera source contains absolute media paths")
+    scrubbed = out / REPOS["camera_traps"]
+    for name in ("camera_image_acquisition_arrivals.csv", "camera_operational_occurrence_arrivals.csv"):
+        atomic_write(scrubbed / "analysis" / name, lambda t, n=name: scrub_csv_copy(src / "analysis" / n, t))
+    atomic_write(scrubbed / "metadata/provenance/ANNOTATION_PROVENANCE.md",
+                 lambda t: scrub_text_copy(src / "metadata/provenance/ANNOTATION_PROVENANCE.md", t,
+                                           extra=[("# ICDE camera-trap annotation provenance", "# Camera-trap annotation provenance")]))
+    files = {
+        "data/events.csv": src / "data/events.csv",
+        "data/occurrences.csv": src / "data/occurrences.csv",
+        "data/images.csv": src / "data/images.csv",
+        "media_manifest.csv": src / "media_manifest.csv",
+        "analysis/camera_arrival_statistics.json": src / "analysis/camera_arrival_statistics.json",
+        "metadata/DATA_DICTIONARY.md": src / "metadata/DATA_DICTIONARY.md",
+        "metadata/taxonomy_mapping.csv": src / "metadata/taxonomy_mapping.csv",
+        "metadata/deployments.csv": camera_deployments(tables, images),
+        "metadata/camera_models.csv": tables["Camera Trap Configurations / Camera Models"],
+        "metadata/camera_geolocation_config.csv": SNAP / "scripts/georeferencing/camera_geolocation_config.csv",
+        "metadata/timestamp_summary_by_camera.csv": TS_MAP / "camera_timestamp_summary_by_camera.csv",
+        "metadata/timestamp_summary_by_date.csv": TS_MAP / "camera_timestamp_summary_by_date.csv",
+        "metadata/provenance/tw07_ct05_sensitive_media_disposition_20261008.json":
+            SNAP / "tw07_ct05_sensitive_media_disposition_20261008.json",
+        "scripts/geolocate_camera_trap_treeoflife_predictions.py":
+            PUBLIC_SCRIPTS / "georeferencing/geolocate_camera_trap_treeoflife_predictions.py",
+        "scripts/generate_darwin_core.py": PUBLIC_SCRIPTS / "darwin_core/generate_darwin_core.py",
+    }
+    return files, {
+        "events": len(ev), "occurrences": len(oc), "images_total": len(images),
+        "all_geometry_provenance_is_SAM3": True,
+        "absolute_media_paths": 0,
+        "water_association_attributes_absent": True,
+        "pass": True,
+    }
+
+
+# --------------------------------------------------------------------------
+# Bioacoustics
+# --------------------------------------------------------------------------
+def build_bioacoustics(out, tables):
+    ev = read(SNAP / "bioacoustics_complete/darwin_core_events_bioacoustics_complete.csv")
+    oc = read(SNAP / "bioacoustics_complete/darwin_core_occurrences_bioacoustics_complete.csv")
+    rec = read(SNAP / "bioacoustics_complete/RAW_WAV_RECONCILIATION_20261008.csv")
+    rec.insert(0, "release_path", rec.raw_wav_path.map(lambda p: rel_media(PROJECT_PREFIX + p)))
+    included = rec.release_disposition.str.startswith("included")
+    path_by_event = rec[included].set_index("darwin_core_eventID").release_path
+    # associatedMedia is a bare WAV filename; resolve it through the Event.
+    oc["associatedMedia"] = oc.eventID.map(path_by_event).fillna(oc.associatedMedia)
+    ev["associatedMedia"] = ev.eventID.map(path_by_event).fillna("")
+
+    disposition = dict(zip(rec.release_path, rec.release_disposition))
+    def excl(rel):
+        d = disposition.get(rel)
+        return None if d is None or d.startswith("included") else d.split(";")[0] + ":" + rec.loc[
+            rec.release_path == rel, "reconciliation_status"].iloc[0]
+    manifest = media_inventory("bioacoustics", excl)
+
+    t = {k: v for k, v in tables.items() if k.startswith("Bioacoustic Recorder Configurations / Campaign")}
+    dep = pd.concat([v.assign(campaign=("thewilds_20250630_20250703" if "June" in k else "thewilds_20250804_20250828"))
+                     for k, v in t.items()], ignore_index=True)
+    dep.insert(0, "deploymentID", dep.Site.str.extract(r"^(TW\d+)")[0] + "-" + dep.Recorder)
+    dep.insert(1, "model", "Wildlife Acoustics Song Meter Micro 2")
+    counts = rec.groupby(["source_campaign_directory", "deploymentID"]).release_disposition.agg(
+        wav_paths="size", wav_released=lambda s: s.str.startswith("included").sum()).reset_index()
+    counts["campaign"] = counts.source_campaign_directory.map(
+        {"campaign1": "thewilds_20250630_20250703", "campaign2": "thewilds_20250804_20250828"})
+    dep = dep.merge(counts.drop(columns="source_campaign_directory"), on=["campaign", "deploymentID"], how="left")
+    birdnet = pd.DataFrame([{"software": "BirdNET Analyzer", "version": "2.4", "segment_length_s": 3.0,
+                             "sample_rate_hz": 48000, "bandpass_hz": "0-15000",
+                             "occurrence_unit": "one row per recording x species; individualCount = detections"}])
+    ev, oc = strip_cluster_paths(ev), strip_cluster_paths(oc)
+    res = check_tables("bioacoustics", ev, oc, manifest, extra_event_ids={"TheWilds_2025"})
+    files = {
+        "data/events.csv": ev, "data/occurrences.csv": oc, "data/recordings.csv": rec,
+        "media_manifest.csv": manifest,
+        "metadata/deployments.csv": dep, "metadata/birdnet_settings.csv": birdnet,
+        "metadata/provenance/RAW_WAV_RECONCILIATION_20261008.md":
+            SNAP / "bioacoustics_complete/RAW_WAV_RECONCILIATION_20261008.md",
+        "metadata/provenance/bioacoustic_event_repair_20261008.json": SNAP / "bioacoustic_event_repair_20261008.json",
+        "scripts/reconcile_bioacoustic_raw_wavs.py": PUBLIC_SCRIPTS / "darwin_core/reconcile_bioacoustic_raw_wavs.py",
+    }
+    return files, res
+
+
+# --------------------------------------------------------------------------
+# Drones
+# --------------------------------------------------------------------------
+def build_drones(out, tables):
+    ev = read(SNAP / "drones/darwin_core_events_drones.csv")
+    oc = read(SNAP / "drones/darwin_core_occurrences_drones.csv")
+    manifest = media_inventory("drones")
+    videos = manifest[manifest.release_path.str.lower().str.endswith((".mp4", ".mov"))]
+    by_stem = {}
+    for p in videos.release_path:
+        flight_dir = p.split("/")[2]
+        by_stem.setdefault((flight_dir, Path(p).stem), []).append(p)
+
+    # Defect 21: the manual starling supplement was appended twice. Drop exact
+    # duplicate rows so every occurrenceID is unique.
+    before = len(oc)
+    oc = oc.drop_duplicates().reset_index(drop=True)
+    rep = {"exact_duplicate_rows_dropped": before - len(oc),
+           "duplicate_occurrenceIDs_remaining": int(oc.occurrenceID.duplicated().sum())}
+    # Frame images (purged scratch frames, or frames exported to an external
+    # annotation workspace) are not shipped; fall back to the retained video.
+    frame_re = re.compile(r"/drones/([^/]+)/.*?/([^/]+)_frames/[^/]+$")
+    def video_for(frame):
+        m = frame_re.search(frame)
+        if not m:
+            return None
+        hits = by_stem.get((m.group(1), m.group(2)), [])
+        return hits[0] if len(hits) == 1 else None
+    frames = oc.associatedMedia.str.startswith("/") & oc.associatedMedia.map(lambda p: rel_media(p) is None)
+    fallback = oc.loc[frames, "associatedMedia"].map(video_for)
+    rep.update({"frame_rows": int(frames.sum()),
+                "frame_rows_unique": int(oc.loc[frames, "associatedMedia"].nunique()),
+                "frame_rows_unmapped": int(fallback.isna().sum())})
+    ok = fallback.dropna().index
+    oc.loc[ok, "associatedMedia"] = fallback[ok]
+    oc.loc[ok, "occurrenceRemarks"] = oc.loc[ok, "occurrenceRemarks"] + \
+        "; associated_media_source=original_video_fallback_frame_not_distributed"
+    oc["associatedMedia"] = oc.associatedMedia.map(lambda p: rel_media(p) or p)
+    # Per-detection table for frame extraction from the released raw videos.
+    # detection_id = <video>_<frame:06d>_<k>; frame is the native frame index
+    # (frames were sampled every 15th frame for SAM3).
+    det = pd.DataFrame({
+        "occurrenceID": oc.occurrenceID, "eventID": oc.eventID, "video_path": oc.associatedMedia,
+        "detection_id": oc.occurrenceRemarks.str.extract(r"detection_id=([^;]+)")[0]})
+    det["frame_index"] = det.detection_id.str.extract(r"_(\d{6})_\d+$")[0].astype("Int64")
+    bbox = oc.identificationRemarks.str.extract(r"bbox xywh\(px\): \[([^\]]+)\]")[0].str.split(",", expand=True)
+    for i, c in enumerate(["bbox_x", "bbox_y", "bbox_w", "bbox_h"]):
+        det[c] = pd.to_numeric(bbox[i]).round(2)
+    det["scientificName"], det["vernacularName"] = oc.scientificName, oc.vernacularName
+    det["top1_score"] = pd.to_numeric(oc.identificationRemarks.str.extract(r"top1 score: ([0-9.eE-]+)")[0])
+    det["georef_status"] = oc.occurrenceRemarks.str.extract(r"drone_georef_status=([^;]+)")[0]
+    det = det[det.detection_id.notna()]  # the 644 curated review records carry no frame id
+    rep["detections_with_frame_index"] = int(det.frame_index.notna().sum())
+    # Remaining frame PNGs that exist on disk are derived products, not raw media.
+    flights = ev[ev.eventType == "SensorDeployment"][["eventID", "eventDate", "locationID", "fieldNotes"]].copy()
+    flights["platform"] = flights.locationID.map(
+        lambda s: "DJI Air 2" if "djiair" in s else "DJI (unspecified airframe)" if "DJI" in s else "Parrot Anafi 4K")
+    media_events = ev[ev.eventType == "MediaEvent"]
+    flights["videos"] = flights.eventID.map(media_events.parentEventID.value_counts()).fillna(0).astype(int)
+    flights["occurrences"] = flights.eventID.map(
+        oc.eventID.map(media_events.set_index("eventID").parentEventID).value_counts()).fillna(0).astype(int)
+    ev, oc = strip_cluster_paths(ev), strip_cluster_paths(oc)
+    res = check_tables("drones", ev, oc, manifest, extra_event_ids={"TheWilds_2025"})
+    res.update(rep)
+    res["pass"] = res["pass"] and rep["duplicate_occurrenceIDs_remaining"] == 0
+    files = {
+        "data/events.csv": ev, "data/occurrences.csv": oc, "data/detections.csv": det,
+        "media_manifest.csv": manifest,
+        "analysis/bird_water_association_per_box.csv": strip_cluster_paths(read(BIRD_WATER)),
+        "metadata/flights.csv": flights, "metadata/platforms.csv": tables["Drone Configurations"],
+        "metadata/drone_projection_config.csv": SNAP / "scripts/georeferencing/drone_projection_config.csv",
+        "metadata/provenance/invalid_drone_date_repair_20261008.json": SNAP / "invalid_drone_date_repair_20261008.json",
+        "metadata/provenance/invalid_drone_coordinate_repair_20261009.json":
+            SNAP / "invalid_drone_coordinate_repair_20261009.json",
+        "metadata/provenance/drone_missing_media_repair_20261010.json": SNAP / "drone_missing_media_repair_20261010.json",
+        "metadata/provenance/excluded_drone_media_event_repair_20261008.json":
+            SNAP / "excluded_drone_media_event_repair_20261008.json",
+        "scripts/geolocate_drone_treeoflife_predictions.py":
+            PUBLIC_SCRIPTS / "georeferencing/geolocate_drone_treeoflife_predictions.py",
+        "scripts/generate_darwin_core.py": PUBLIC_SCRIPTS / "darwin_core/generate_darwin_core.py",
+    }
+    return files, res
+
+
+# --------------------------------------------------------------------------
+# Citizen science and umbrella (need files that are currently owner-only)
+# --------------------------------------------------------------------------
+def build_citizen_science(out, tables):
+    evp = SNAP / "citizen_science/darwin_core_events_citizen_science.csv"
+    ocp = SNAP / "citizen_science/darwin_core_occurrences_citizen_science.csv"
+    if not (readable(evp) and readable(ocp)):
+        raise PermissionError(f"unreadable: {evp.name}, {ocp.name} (owner-only permissions)")
+    ev, oc = read(evp), read(ocp)
+    oc["associatedMedia"] = oc.associatedMedia.map(lambda p: rel_media(p) or p)
+    # The Sep 8 Google Photos album export is not in the project archive, so its
+    # 50 photos cannot be distributed: keep the Occurrences, clear the link, say why.
+    album = oc.associatedMedia.str.startswith("Google Photos Takeout/")
+    oc.loc[album, "occurrenceRemarks"] = oc.loc[album, "occurrenceRemarks"] + \
+        "; media_not_distributed=Google Photos album export not retained in project archive (was " + \
+        oc.loc[album, "associatedMedia"] + ")"
+    oc.loc[album, "associatedMedia"] = ""
+    manifest = media_inventory("citizen_science")
+    ev, oc = strip_cluster_paths(ev), strip_cluster_paths(oc)
+    res = check_tables("citizen_science", ev, oc, manifest, extra_event_ids={"TheWilds_2025"})
+    res["occurrences_media_not_distributed"] = int(album.sum())
+    return {"data/events.csv": ev, "data/occurrences.csv": oc, "media_manifest.csv": manifest,
+            "metadata/provenance/citizen_science_media_reconciliation_20261008.json":
+                SNAP / "citizen_science_media_reconciliation_20261008.json"}, res
+
+
+def build_core(out, tables):
+    """Study-level dataset: validated merged tables, text records, sensor
+    configurations, supplement, defect inventory, and the arrival traces."""
+    src = CAMERA_SOURCE / "smartwilds/data"
+    ev = read(src / "events_all_modalities.csv")
+    oc = read(src / "occurrences_all_modalities.csv")
+    if len(ev) != 22242 or len(oc) != 401602 or oc.occurrenceID.duplicated().any():
+        raise RuntimeError("validated merged-table integrity check failed")
+    # The merged table carries the drone and citizen-science rows with their
+    # archive media paths. Take the release-relative values from the staged
+    # modality tables (same occurrenceID), then rewrite anything left over.
+    rep = {}
+    for repo in ("smartwilds-drones", "smartwilds-citizen-science"):
+        p = out / repo / "data/occurrences.csv"
+        if not p.exists():
+            rep[f"{repo}_not_staged"] = True
+            continue
+        m = read(p).drop_duplicates("occurrenceID").set_index("occurrenceID")
+        idx = oc.occurrenceID.isin(m.index)
+        for c in ("associatedMedia", "occurrenceRemarks", "associatedReferences"):
+            if c in m.columns and c in oc.columns:
+                oc.loc[idx, c] = oc.loc[idx, "occurrenceID"].map(m[c]).values
+        rep[f"{repo}_rows_aligned"] = int(idx.sum())
+    oc["associatedMedia"] = oc.associatedMedia.map(lambda p: rel_media(p) or p)
+    ev, oc = strip_cluster_paths(ev), strip_cluster_paths(oc)
+    res = {"events": len(ev), "occurrences": len(oc),
+           "absolute_media_paths": int(oc.associatedMedia.str.startswith("/").sum()),
+           "cells_with_cluster_paths": count_cluster_paths(ev) + count_cluster_paths(oc)}
+    res.update(rep)
+    res["pass"] = res["absolute_media_paths"] == 0 and res["cells_with_cluster_paths"] == 0
+    files = {
+        "data/events_all_modalities.csv": ev,
+        "data/occurrences_all_modalities.csv": oc,
+        "data/text_events.csv": SNAP / "text_data/darwin_core_events_text.csv",
+        "data/text_occurrences.csv": SNAP / "text_data/darwin_core_occurrences_text.csv",
+        "data/site_selection_rationale.csv": SNAP / "text_data/site_selection_rationale.csv",
+        "metadata/sensor_configurations.md": SENSOR_MD,
+        "SUPPLEMENTAL.md": CARDS / "SUPPLEMENTAL.md",
+        "metadata/DEFECT_INVENTORY.md": CARDS / "DEFECT_INVENTORY.md",
+        "metadata/defect_inventory.csv": CARDS / "defect_inventory.csv",
+        "scripts/refresh_camera_arrival_metadata.py": ROOT / "release/scripts/refresh_camera_arrival_metadata.py",
+        "scripts/generate_arrival_traces_v2.py": PUBLIC_SCRIPTS / "arrival_traces/generate_arrival_traces_v2.py",
+    }
+    # Non-camera arrival traces come from the integration pass; the camera
+    # streams are refreshed afterwards from the validated derivative. The two
+    # notes about the first-generation camera trace no longer apply.
+    atomic_write(out / REPOS["core"] / "metadata/provenance/DARWIN_CORE_AUDIT_20261010_FINAL.json",
+                 lambda t: scrub_text_copy(SNAP / "DARWIN_CORE_AUDIT_20261010_FINAL.json", t))
+    skip = {"DEPRECATED_LEGACY_PRIMITIVES.md", "UNAVAILABLE_CAMERA_IMAGE_CAPTURE_ACQUISITION.md"}
+    drone_media = None
+    for p in sorted((SNAP / "analysis/arrival_traces_v2").iterdir()):
+        if p.name in skip or p.name.startswith("camera_"):
+            continue
+        if p.suffix == ".csv":
+            tr = read(p)
+            if tr.associatedMedia.str.startswith("/").any():
+                if drone_media is None:
+                    drone_media = oc.drop_duplicates("occurrenceID").set_index("occurrenceID").associatedMedia
+                by_source = tr.sourceID.map(drone_media)
+                tr["associatedMedia"] = by_source.where(by_source.notna(), tr.associatedMedia.map(lambda x: rel_media(x) or x))
+            tr = strip_cluster_paths(tr)
+            if "sourceSnapshot" in tr.columns:
+                tr["sourceSnapshot"] = tr.sourceSnapshot.str.replace(SNAP.name, "integration_pass_20261022", regex=False)
+            res[f"arrival_trace_cluster_paths:{p.name}"] = count_cluster_paths(tr)
+            res["pass"] = res["pass"] and res[f"arrival_trace_cluster_paths:{p.name}"] == 0
+            files[f"analysis/arrival_traces_v2/{p.name}"] = tr
+        else:
+            files[f"analysis/arrival_traces_v2/{p.name}"] = p
+    return files, res
+
+
+def main():
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--out", default=str(RELEASE_DIR))
+    ap.add_argument("--only", nargs="*", choices=list(REPOS))
+    args = ap.parse_args()
+    out = Path(args.out)
+    out.mkdir(parents=True, exist_ok=True)
+    tables = parse_md_tables(SENSOR_MD)
+    audit["built_utc"] = datetime.now(timezone.utc).isoformat(timespec="seconds")
+    builders = {"camera_traps": build_camera, "bioacoustics": build_bioacoustics, "drones": build_drones,
+                "citizen_science": build_citizen_science, "core": build_core}
+    for key, fn in builders.items():
+        if args.only and key not in args.only:
+            continue
+        repo = REPOS[key]
+        print(f"[{repo}] building ...", flush=True)
+        for rel in STALE:
+            (out / repo / rel).unlink(missing_ok=True)
+        card = CARDS / f"{repo}.md"
+        if card.exists():
+            atomic_write(out / repo / "README.md", lambda t: shutil.copyfile(card, t))
+        imgs = CARDS / "images" / repo
+        if imgs.is_dir():
+            # Release-card images are static assets. Avoid rewriting existing
+            # owner-managed media assets during an in-place table rebuild.
+            target_images = out / repo / "images"
+            if not target_images.exists():
+                shutil.copytree(imgs, target_images)
+        try:
+            files, res = fn(out, tables)
+        except PermissionError as e:
+            audit["blocked"].append({"repo": repo, "reason": str(e)})
+            print(f"[{repo}] BLOCKED: {e}", flush=True)
+            continue
+        files[f"scripts/{Path(__file__).name}"] = Path(__file__)
+        d = write_repo(out, repo, files)
+        if key != "camera_traps":
+            write_dictionary(d, repo)
+        write_checksums(d)
+        audit["repos"][repo] = res
+        print(f"[{repo}] {json.dumps({k: v for k, v in res.items() if not isinstance(v, (list, dict))})}",
+              flush=True)
+    # The camera arrival streams in the study-level dataset come from the
+    # validated camera derivative; refresh them once both parts are staged.
+    requested = set(args.only or REPOS)
+    if {"camera_traps", "core"}.issubset(requested):
+        subprocess.run([sys.executable, str(ROOT / "release/scripts/refresh_camera_arrival_metadata.py"),
+                        "--release", str(out), "--snapshot", str(SNAP), "--source", str(CAMERA_SOURCE)], check=True)
+        write_dictionary(out / REPOS["core"], REPOS["core"])
+        write_checksums(out / REPOS["core"])
+    (out / "build_hf_release.py").unlink(missing_ok=True)  # stray copy from an earlier build
+    (out / "RELEASE_BUILD_AUDIT.json").write_text(json.dumps(audit, indent=2, default=str) + "\n")
+    return 0 if not audit["errors"] else 1
+
+
+if __name__ == "__main__":
+    sys.exit(main())

smartwilds-bioacoustics/scripts/reconcile_bioacoustic_raw_wavs.py

@@ -0,0 +1,214 @@
+#!/usr/bin/env python3
+'''Reconcile raw The Wilds WAV files against corrected Darwin Core Events.'''
+from __future__ import annotations
+
+import csv
+import json
+import os
+import re
+import wave
+from collections import Counter
+from datetime import date
+from pathlib import Path
+
+BASE_DIR = Path(os.environ.get('SMARTWILDS_ARCHIVE', '.')).resolve()
+SNAPSHOT_DIR = Path(os.environ.get('SMARTWILDS_DWC_OUT', BASE_DIR / 'processing_output/darwin_core_release'))
+OUT_DIR = SNAPSHOT_DIR / 'bioacoustics_complete'
+OUT_CSV = OUT_DIR / 'RAW_WAV_RECONCILIATION_20261008.csv'
+OUT_NOTE = OUT_DIR / 'RAW_WAV_RECONCILIATION_20261008.md'
+OUT_AUDIT = SNAPSHOT_DIR / 'bioacoustic_event_repair_20261008.json'
+EVENTS = OUT_DIR / 'darwin_core_events_bioacoustics_complete.csv'
+ROOTS = [
+    (BASE_DIR / 'thewilds_20250630_20250703' / 'bioacoustics', date(2025, 6, 30), date(2025, 7, 3), 'campaign1'),
+    (BASE_DIR / 'thewilds_20250804_20250828' / 'bioacoustics', date(2025, 8, 4), date(2025, 8, 28), 'campaign2'),
+]
+RESULT_DIRS = [
+    BASE_DIR / 'processing_output/detections/acoustics/BirdNet_EveryHour',
+    BASE_DIR / 'BioacousticAnalysis/BirdnetCSV_SunriseSunset',
+]
+NAME_RE = re.compile(r'^(?:[A-Z0-9]+)_(\d{8})_(\d{6})$', re.I)
+EVENT_RE = re.compile(r'Recording: ([^;]+)\.wav; Recorder: ([^;]+);')
+FIELDS = [
+    'raw_wav_path', 'source_campaign_directory', 'deploymentID', 'recording_file',
+    'recording_date', 'recording_time', 'in_source_campaign_window', 'wav_bytes',
+    'wav_header_status', 'duration_seconds', 'birdnet_result_path',
+    'birdnet_result_status', 'darwin_core_eventID', 'reconciliation_status',
+    'release_disposition',
+]
+
+
+def deployment(path: Path) -> tuple[str, str]:
+    for part in path.parts:
+        match = re.fullmatch(r'(TW\d+)-(SM\d+)', part)
+        if match:
+            return match.group(1), match.group(2)
+    raise ValueError(f'No deployment in {path}')
+
+
+def wave_details(path: Path) -> tuple[str, str]:
+    try:
+        with wave.open(str(path), 'rb') as audio:
+            duration = audio.getnframes() / audio.getframerate()
+            return 'readable', f'{duration:.6f}'
+    except Exception as exc:
+        return f'unreadable:{type(exc).__name__}', ''
+
+
+def result_status(path: Path | None) -> str:
+    if path is None:
+        return 'absent'
+    with path.open(newline='') as handle:
+        reader = csv.DictReader(handle)
+        return 'header_only_zero_detection' if next(reader, None) is None else 'nonempty'
+
+
+def main() -> None:
+    results = {}
+    for directory in RESULT_DIRS:
+        for path in directory.glob('*.BirdNET.results.csv'):
+            results[path.name.removesuffix('.BirdNET.results.csv')] = path
+
+    events = {}
+    with EVENTS.open(newline='') as handle:
+        for row in csv.DictReader(handle):
+            if row['eventType'] != 'AudioRecording':
+                continue
+            match = EVENT_RE.search(row['fieldNotes'])
+            if not match:
+                raise ValueError(f'Unparseable recording provenance: {row["eventID"]}')
+            recording, recorder = match.groups()
+            key = (row['locationID'], recorder, recording.split('_', 1)[1])
+            if key in events:
+                raise ValueError(f'Duplicate release recording key: {key}')
+            events[key] = (row['eventID'], 'header-only BirdNET result' in row['fieldNotes'])
+
+    rows = []
+    for root, start, end, campaign in ROOTS:
+        for path in sorted(root.rglob('*')):
+            if not (path.is_file() and path.suffix.lower() == '.wav'):
+                continue
+            name_match = NAME_RE.match(path.stem)
+            if not name_match:
+                raise ValueError(f'Unparseable raw WAV name: {path.name}')
+            site, recorder = deployment(path)
+            ymd, hms = name_match.groups()
+            recording_date = date.fromisoformat(f'{ymd[:4]}-{ymd[4:6]}-{ymd[6:]}')
+            in_window = start <= recording_date <= end
+            event_entry = events.get((site, recorder, f'{ymd}_{hms}'))
+            event_id = event_entry[0] if event_entry else ''
+            is_header_only = event_entry[1] if event_entry else False
+            result = results.get(path.stem)
+            if not in_window:
+                status = 'transfer_duplicate'
+                disposition = 'excluded_from_release; canonical recording is represented under its original campaign/deployment'
+            elif is_header_only:
+                status = 'processed_zero_detection'
+                disposition = 'included_as_AudioRecording_Event_without_Occurrences'
+            elif event_id:
+                status = 'represented_recording'
+                disposition = 'included_as_AudioRecording_Event'
+            elif result is None:
+                status = 'truncated_or_corrupt_unprocessed'
+                disposition = 'excluded_from_release; no BirdNET result and insufficient/corrupt source audio'
+            else:
+                raise ValueError(f'Unexpected unmatched processed recording: {path}')
+            if status == 'truncated_or_corrupt_unprocessed':
+                header_status, duration = wave_details(path)
+            else:
+                header_status, duration = 'not_checked_release_represented_or_transfer', ''
+            rows.append({
+                'raw_wav_path': str(path.relative_to(BASE_DIR)),
+                'source_campaign_directory': campaign,
+                'deploymentID': f'{site}-{recorder}',
+                'recording_file': path.name,
+                'recording_date': recording_date.isoformat(),
+                'recording_time': f'{hms[:2]}:{hms[2:4]}:{hms[4:]}',
+                'in_source_campaign_window': str(in_window).lower(),
+                'wav_bytes': path.stat().st_size,
+                'wav_header_status': header_status,
+                'duration_seconds': duration,
+                'birdnet_result_path': str(result.relative_to(BASE_DIR)) if result else '',
+                'birdnet_result_status': ('header_only_zero_detection' if is_header_only else ('nonempty' if result else 'absent')),
+                'darwin_core_eventID': event_id,
+                'reconciliation_status': status,
+                'release_disposition': disposition,
+            })
+
+    counts = Counter(row['reconciliation_status'] for row in rows)
+    expected = Counter({
+        'represented_recording': 2307,
+        'processed_zero_detection': 64,
+        'transfer_duplicate': 78,
+        'truncated_or_corrupt_unprocessed': 4,
+    })
+    if counts != expected:
+        raise ValueError(f'Unexpected reconciliation counts: {counts}')
+    if len(events) != 2371:
+        raise ValueError(f'Expected 2,371 AudioRecording Events, found {len(events)}')
+
+    with OUT_CSV.open('w', newline='', encoding='utf-8') as handle:
+        writer = csv.DictWriter(handle, fieldnames=FIELDS)
+        writer.writeheader()
+        writer.writerows(rows)
+
+    note = f'''# Raw WAV Reconciliation
+
+## Outcome
+
+The corrected release has **2,371 unique `AudioRecording` Events**. It contains
+no duplicate Event for transferred raw media and keeps the existing 9,960
+BirdNET Occurrences unchanged.
+
+| Raw-file disposition | Files | Release treatment |
+| --- | ---: | --- |
+| Represented recording | {counts['represented_recording']} | One `AudioRecording` Event |
+| Processed, zero detections | {counts['processed_zero_detection']} | One `AudioRecording` Event; no Occurrence |
+| Transfer duplicate | {counts['transfer_duplicate']} | Excluded; source belongs to another campaign/deployment |
+| Truncated/corrupt, unprocessed | {counts['truncated_or_corrupt_unprocessed']} | Excluded; no BirdNET result |
+| **Raw files on disk** | **{len(rows)}** | **Reconciled** |
+
+## Why The Earlier Count Was Misleading
+
+The former comparison of 2,453 raw WAV files with 2,307 Events treated every
+physical file path as a separate capture. Seventy-eight Campaign-1 WAVs were
+transferred into the Campaign-2 `TW02-SM04/Data` directory; they are duplicate
+transfers, not new Campaign-2 recordings. The remaining 68 unmatched
+in-window files divide into 64 successfully processed files with header-only
+BirdNET results (zero detections) and four truncated/corrupt files with no
+BirdNET result.
+
+The 64 zero-detection recordings are now represented as Events with a field
+note identifying the header-only BirdNET result. They deliberately generate no
+Occurrence. The four failures remain out of the Event table and are listed in
+the accompanying CSV with their header status and duration.
+
+## Integrity Checks
+
+- Every `AudioRecording` Event has a unique recording key: deployment, recorder,
+  date, and time.
+- The 78 transfer duplicates never create a second Event.
+- All 9,960 Occurrences retain valid Event links.
+- The CSV provides one row per raw WAV path, including path, deployment, WAV
+  header status, duration, BirdNET-result status, final Event ID, and release
+  disposition.
+'''
+    OUT_NOTE.write_text(note, encoding='utf-8')
+    audit = {
+        'repair': 'bioacoustic AudioRecording completeness and raw-WAV reconciliation',
+        'snapshot': SNAPSHOT_DIR.name,
+        'raw_wav_files': len(rows),
+        'audio_recording_events': len(events),
+        'occurrences_unchanged': 9960,
+        'reconciliation_status_counts': dict(sorted(counts.items())),
+        'release_policy': {
+            'transfer_duplicate': 'excluded; no duplicate Event',
+            'processed_zero_detection': 'included as Event with no Occurrence',
+            'truncated_or_corrupt_unprocessed': 'excluded; documented in reconciliation CSV',
+        },
+    }
+    OUT_AUDIT.write_text(json.dumps(audit, indent=2) + '\n', encoding='utf-8')
+    print(json.dumps(audit, indent=2))
+
+
+if __name__ == '__main__':
+    main()

smartwilds-camera-traps/CHECKSUMS.sha256

@@ -0,0 +1,19 @@
+b2313d24026328b451ff7decdcb56e6d8becd6727a7a7ff619a378d9913dc287  analysis/camera_arrival_statistics.json
+d959c270284e11a477b3b6d6452b4e390a775f1ff2c5e2222852a2442f74d1f5  analysis/camera_image_acquisition_arrivals.csv
+fcd8bf4cbf4c39f51e2ec80396d25560e6e70066282931475d9f50145ea78281  analysis/camera_operational_occurrence_arrivals.csv
+3d0641255c3e5ed0be27548c526c1627e61bba40076321424846958486183c8b  data/events.csv
+c445933a1aa68b43cb1d5f73d94831e4e9c2d9b6367f5bf78fd1da35355710f9  data/images.csv
+f777c5eb279af3306dad1feec8448d3ae8768b849108cbb1ed54d58e433f62d2  data/occurrences.csv
+c8cf498a546e4f4fe3015c204d9d830174197054312cd373775b9ffbf916f523  media_manifest.csv
+01aa5c3f54c8ecc215126d889851907659b2a8185578022e35b49ad0b1240276  metadata/DATA_DICTIONARY.md
+7410a9b0a876ddce34d4b0e46a01ec07d0a14c46329d8759263ff3eb6e6a148f  metadata/camera_geolocation_config.csv
+03207450d21d8b705d08c66bb2f3d541bc9c57c0ae49d417248f3b5970ca4078  metadata/camera_models.csv
+cd22b797f1d4ec7ac3a688c6378a169889b73ff5000d81d6e09bc51b3999ad3a  metadata/deployments.csv
+ebf8f50f331479e0ff3ed478154f96af926cca25fe0d8b1516fb593d1dedac88  metadata/provenance/ANNOTATION_PROVENANCE.md
+4c8595701d0f0e3ad59432389e28b52706243bd883be78628b240e49f3d6555a  metadata/provenance/tw07_ct05_sensitive_media_disposition_20261008.json
+b0a4165679cc141eed45e8e36da9f2c821e3626393191323b0b2effd91cd4a43  metadata/taxonomy_mapping.csv
+a0801f8542b2bb9cd1a099a1402046aa8c7f3253fe9dfdf83f3352f0e88a7825  metadata/timestamp_summary_by_camera.csv
+66ca26fb7d9045201c670a914d210f5ec40ace34962261a611368862fbc5896c  metadata/timestamp_summary_by_date.csv
+f2efc6692a5d7add094933b71b4f64a15efa2fd1c2578bb66620ff161747934a  scripts/build_hf_release.py
+6e90073a9d7bcaf1c6d37fbede20d7444d07792000888c8d729e5787986d9545  scripts/generate_darwin_core.py
+fcbd9ad5b36b718118981ca983349ceb986edec8d1e0502072745560661a5e26  scripts/geolocate_camera_trap_treeoflife_predictions.py

smartwilds-camera-traps/README.md

@@ -0,0 +1,227 @@
+---
+license: cc0-1.0
+language:
+- en
+pretty_name: SmartWilds Camera Traps (The Wilds 2025)
+task_categories:
+- object-detection
+- image-classification
+tags:
+- biology
+- ecology
+- biodiversity
+- camera-trap
+- darwin-core
+- smartwilds
+- the-wilds
+- ohio
+size_categories:
+- 100K<n<1M
+description: >-
+  235,778 processed camera-trap images from The Wilds conservation center
+  (Ohio, USA), June-August 2025. The release contains 72,775 retained animal
+  bounding boxes with original SAM3 geometry, labels from Resolve–Clear–Defer
+  prototype triage and targeted human taxonomic corrections, and 19,785 Darwin
+  Core Events.
+
+configs:
+- config_name: occurrences
+  data_files: data/occurrences.csv
+- config_name: events
+  data_files: data/events.csv
+- config_name: images
+  data_files: data/images.csv
+---
+
+# Dataset Card for SmartWilds Camera Traps
+
+This camera-trap component of **SmartWilds** contains the complete processed
+image inventory collected at The Wilds in summer 2025, corrected image
+timestamps, deployment metadata, and a curated operational animal-detection
+table.
+
+## Dataset Details
+
+| Resource | Count | Description |
+|---|---:|---|
+| Processed image inventory | 235,778 | All valid processed camera images, including images without retained animal boxes. |
+| Camera Events | 19,785 | One Project, 12 SensorDeployments, and 19,772 per-image Events. |
+| Retained operational Occurrences | 72,775 | Animal bounding boxes, each with original SAM3 proposal geometry. |
+| Images with retained Occurrences | 19,726 | Images referenced by one or more retained operational boxes. |
+| Zero-occurrence Image Events | 46 | Legitimate Image Events with no retained operational box. |
+| Curated images approved for upload after human privacy review | 31 | Screened candidates human-reviewed and approved for upload. |
+
+The Event hierarchy is rooted at `TheWilds_2025`. The camera Project is
+`SmartWilds_camera_traps_curated`; its 12 SensorDeployments and 19,772 Image
+Events preserve the final curated hierarchy. The image inventory is broader
+than the curated Image Event subtree, so an image with `occurrence_count = 0`
+does not establish animal absence.
+
+### Files
+
+```text
+smartwilds-camera-traps/
+├── data/
+│   ├── events.csv                 # Darwin Core Event hierarchy
+│   ├── occurrences.csv            # 72,775 retained animal boxes
+│   └── images.csv                 # 235,778 processed images and corrected times
+├── analysis/                      # acquisition and operational-inference traces
+├── metadata/
+│   ├── DATA_DICTIONARY.md
+│   ├── taxonomy_mapping.csv
+│   ├── deployments.csv            # per-deployment sensor settings and geolocation config
+│   ├── camera_models.csv
+│   ├── timestamp_summary_by_camera.csv, timestamp_summary_by_date.csv
+│   └── provenance/                # annotation and release-integration records
+├── scripts/                       # Darwin Core generator, georeferencing, release build
+├── media_manifest.csv             # relative raw-media packaging references
+└── CHECKSUMS.sha256
+```
+
+All `associatedMedia` and manifest `release_path` values are release-relative
+`media/...` paths. The media manifest references raw media; it does not copy
+media into this release tree.
+
+### Supported Tasks and Applications
+
+This dataset supports ecological analysis, data-quality research, and sensing-systems research; the computer-vision tasks apply once the images are released.
+
+**🌿 Ecological Applications:**
+- Species occurrence and relative activity by site, hour, and date from 72,775 operational detections
+- Diel activity patterns from corrected capture timestamps
+- Co-occurrence with the acoustic and drone modalities at shared sites (TW02, TW06, TW09)
+- Detection-rate comparison across camera models, trigger modes, and placements
+
+**🧹 Data-Quality and Metadata Research:**
+- Timestamp reconstruction: 18 per-deployment correction methods with method and confidence per image
+- Label provenance: triage-resolved, human-reviewed, forced nearest-prototype, and human-edited labels kept distinct
+- Semi-automated labeling evaluation against the human-reviewed subset
+
+**⚙️ Sensing-Systems Research:**
+- Image-capture and detection arrival traces for workload characterization (burstiness, diurnal rate changes)
+- Blank-image filtering and storage savings from the full image inventory
+- Edge hardware provisioning for camera-trap detection pipelines
+
+**🤖 Computer Vision Tasks (with the image release):**
+- Object detection (SAM3 proposal geometry, 72,775 boxes)
+- Species classification from detection crops
+- Benchmarking semi-automated labeling against the 2,067-box human reference set
+
+## Dataset Details
+
+- **Curated by:** Jenna M. Kline, Anirudh Potlapally, Elizabeth G. Campolongo
+- **Homepage:** <https://github.com/Imageomics/smartwilds> (documentation: <https://imageomics.github.io/smartwilds/>)
+- **Paper:** SmartWilds: Integrating, Aligning, and Documenting Multi-Modal Field Sensing Data (under review)
+
+## Annotation Method and Provenance
+
+SAM3 generated the original animal proposals. Labels were assigned with the
+Resolve–Clear–Defer prototype-triage protocol described by Potlapally et al.
+(2026), using reviewed support examples to resolve, clear, or defer proposals.
+The retained per-box provenance records the resulting label pathway.
+
+| `labelProvenance` | Retained boxes | Interpretation |
+|---|---:|---|
+| `triage_resolved` | 70,728 | Prototype-triage machine assignment. |
+| `triage_reviewed` | 1,807 | Human-reviewed triage label; 81 were targeted reviews outside numbered rounds, including banteng. |
+| `triage_forced_1nn` | 199 | Small proposal unresolved after guarded triage; machine-assigned from its nearest reviewed prototype without the usual decision safeguards. |
+| `human-edit` | 39 | Targeted human taxonomic correction on an existing SAM3 proposal. |
+| `human-reviewed` | 2 | Targeted human review label on an existing SAM3 proposal. |
+| **Total** | **72,775** | **Retained operational occurrences.** |
+
+The 199 forced 1-NN labels correspond to small proposals remaining unresolved
+after guarded Resolve–Clear–Defer triage. They were assigned the label of their
+nearest reviewed prototype without the usual decision safeguards and were not
+individually expert-verified. Every retained row uses original SAM3 proposal
+geometry.
+
+The taxonomy mapping in `metadata/taxonomy_mapping.csv` records the final
+scientific names and ranks for the 13 project source-label classes. Broad
+labels such as `bird` and `insect` remain broad rather than being generalized
+beyond the evidence available for the proposal.
+
+## Timestamps, Media, and Privacy
+
+`data/images.csv` retains the raw timestamp, corrected timestamp, correction
+method, and correction confidence for the full processed inventory. Camera
+Event and Occurrence dates and times use the corrected image timestamps.
+
+`media_manifest.csv` supplies release-relative paths and packaging status for
+raw media. Privacy fields retain screening evidence and record the
+human-review outcome. Media in which a person can be individually identified are removed from the
+release. People too small to identify (a person box under 80 px in the native
+image, where the face is under about 10 px) are retained; every file with a
+larger person detection was reviewed by a person before release.
+The 31 curated images flagged by screening were reviewed and cleared for
+retention.
+
+## Limitations
+
+- Occurrences are detection boxes, not unique animals; repeated triggers and
+  adjacent images can represent the same individual.
+- The operational labels are semi-automated rather than a fully verified
+  object-detection ground-truth set. The 1,807 rows with
+  `labelProvenance = triage_reviewed` are the human-verified triage subset;
+  machine-derived labels elsewhere in the table should not be used as a fully
+  verified object-detection benchmark.
+- A zero `occurrence_count` is not evidence of animal absence.
+
+## Reproducibility
+
+The camera tables are produced by the post-triage annotation pipeline described
+above and packaged by `scripts/build_hf_release.py`, which validates original
+SAM3 geometry and release-relative media paths before staging. The Darwin Core
+generator and the georeferencing script are in `scripts/`.
+
+## Annotation-protocol reference
+
+The semi-automated labeling workflow follows [Potlapally et al. (2026),
+*Resolve, Clear, or Defer: Training-Free Active Prototype Triage for Ecological
+Detection Pipelines*](https://arxiv.org/abs/2611.01466).
+
+## Licensing
+
+Data: [CC0 1.0](https://creativecommons.org/publicdomain/zero/1.0/). Scripts: MIT.
+
+## Citation
+
+```bibtex
+@misc{smartwilds2025_camera_traps,
+  author    = {Kline, Jenna M. and Potlapally, Anirudh and Campolongo, Elizabeth G.},
+  title     = {SmartWilds Camera Traps (The Wilds 2025)},
+  year      = {2026},
+  publisher = {Hugging Face},
+  note      = {DOI pending the raw data release},
+  url       = {https://huggingface.co/datasets/ABC-Center/smartwilds_metadata/tree/main/smartwilds-camera-traps}
+}
+```
+
+Please also cite the paper:
+
+```bibtex
+@article{kline2026smartwilds,
+  author  = {Kline, Jenna M. and Potlapally, Anirudh and Campolongo, Elizabeth G.},
+  title   = {SmartWilds: Integrating, Aligning, and Documenting Multi-Modal Field Sensing Data},
+  year    = {2026},
+  note    = {Under review}
+}
+```
+
+This dataset card follows the [FAIR² Drones Data Standard](https://imageomics.github.io/fair_drones/)
+(Kline and Campolongo, 2026).
+
+## Acknowledgements
+
+This project is supported by the [Imageomics Institute](https://imageomics.org), which is funded by the US National Science Foundation's Harnessing the Data Revolution (HDR) program under [Award #2118240](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2118240) (Imageomics: A New Frontier of Biological Information Powered by Knowledge-Guided Machine Learning), and by the [AI Institute for Intelligent Cyberinfrastructure with Computational Learning in the Environment (ICICLE)](https://icicle.osu.edu/), which is funded by the US National Science Foundation under [Award #2112606](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2112606).
+
+This work is also supported by the AI and Biodiversity Change (ABC) Global Center. The ABC Global Center is funded by the US National Science Foundation under [Award No. 2330423](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2330423&HistoricalAwards=false) and the Natural Sciences and Engineering Research Council of Canada under [Award No. 585136](https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=782440). This dataset draws on research supported by the Social Sciences and Humanities Research Council.
+
+Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation, Natural Sciences and Engineering Research Council of Canada, or Social Sciences and Humanities Research Council.
+
+Special thanks to Dan Beetem and [The Wilds](https://www.thewilds.org/) in Ohio for providing access to deployment sites and supporting this wildlife monitoring research. We thank the Columbus Zoo and Aquarium for their support in facilitating this project. All data collection, including drone flights, was conducted under the supervision of the Director of Animal Management, with permission from The Wilds Animal Care and Use Committee.
+
+This work used the [Ohio Supercomputer Center](https://www.osc.edu/) (Ohio Supercomputer Center. 1987. Ohio Supercomputer Center. Columbus OH: Ohio Supercomputer Center. <https://ror.org/01apna436>).
+
+## Dataset Card Contact
+
+Jenna M. Kline ([kline.377@osu.edu](mailto:kline.377@osu.edu))

smartwilds-camera-traps/analysis/camera_arrival_statistics.json

@@ -0,0 +1,69 @@
+{
+  "definitions": {
+    "rate": "time-eligible arrivals / fixed 600-hour Campaign 2 window",
+    "interarrival": "successive sorted second-level timestamps, ties retained",
+    "cv": "population SD / mean inter-arrival interval",
+    "burstiness": "B=(CV-1)/(CV+1)"
+  },
+  "acquisition": {
+    "stream_id": "camera_image_acquisition",
+    "source_file": "smartwilds-camera-traps/analysis/camera_image_acquisition_arrivals.csv",
+    "source_sha256": "cf3367ff2c824bc83f0e17fb8008b70e30dc5e251cdb51b8c929f65ef6555493",
+    "counts": {
+      "selected_source_rows": 235778,
+      "unique_source_ids": 235778,
+      "by_campaign": {
+        "campaign1": 15081,
+        "campaign2": 220697
+      },
+      "time_eligible_by_campaign": {
+        "campaign1": 15081,
+        "campaign2": 220697
+      },
+      "timestamp_precision": {
+        "second": 235778
+      }
+    },
+    "campaign2_statistics": {
+      "window_start": "2025-08-04T00:00:00",
+      "window_end_exclusive": "2025-08-29T00:00:00",
+      "window_hours": 600,
+      "time_eligible_arrivals": 220697,
+      "observation_window_rate_per_hour": 367.8283333333333,
+      "interarrival_interval_count": 220696,
+      "mean_interarrival_seconds": 9.407497190705767,
+      "interarrival_cv_population": 2.064767366037759,
+      "burstiness_B": 0.3474219211014141
+    }
+  },
+  "operational_inference": {
+    "source_file": "analysis/camera_operational_occurrence_arrivals.csv",
+    "source_sha256": "0c788434afae99a62ff5c0191aac7ba94c89a333a95966accbbe920c417175d9",
+    "counts": {
+      "selected_source_rows": 72775,
+      "unique_source_ids": 72775,
+      "by_campaign": {
+        "campaign1": 2309,
+        "campaign2": 70466
+      },
+      "time_eligible_by_campaign": {
+        "campaign1": 2309,
+        "campaign2": 70466
+      },
+      "timestamp_precision": {
+        "second": 72775
+      }
+    },
+    "campaign2_statistics": {
+      "window_start": "2025-08-04T00:00:00",
+      "window_end_exclusive": "2025-08-29T00:00:00",
+      "window_hours": 600,
+      "time_eligible_arrivals": 70466,
+      "observation_window_rate_per_hour": 117.44333333333333,
+      "interarrival_interval_count": 70465,
+      "mean_interarrival_seconds": 29.44670403746541,
+      "interarrival_cv_population": 12.343094880504369,
+      "burstiness_B": 0.8501097370654086
+    }
+  }
+}

smartwilds-camera-traps/metadata/DATA_DICTIONARY.md

@@ -0,0 +1,51 @@
+# SmartWilds Camera Traps data dictionary
+
+## `data/occurrences.csv`
+
+One retained animal bounding box per row. The operational table contains
+72,775 occurrences; it is not a fully human-verified object-detection
+benchmark.
+
+| Field(s) | Definition |
+|---|---|
+| `occurrenceID`, `eventID`, `associatedMedia` | Stable occurrence identifier, its per-image Event, and the release-relative `media/...` source image. |
+| `bboxX`, `bboxY`, `bboxWidth`, `bboxHeight` | Bounding box in source-image pixel `xywh` coordinates: top-left x/y and width/height. |
+| `geometryProvenance` | `SAM3` for every retained row; all released geometry is the original SAM3 proposal geometry. |
+| `labelProvenance` | `triage_resolved` (70,728 prototype-triage machine labels); `triage_reviewed` (1,807 human-reviewed triage labels); `triage_forced_1nn` (199 small proposals unresolved after guarded triage and assigned their nearest reviewed-prototype label without usual decision safeguards; machine-assigned, not individually expert-verified); `human-edit` (39 targeted taxonomic corrections); or `human-reviewed` (2 targeted human review labels). |
+| `triageRound` | Numbered Resolve–Clear–Defer review round where applicable. It is blank for 83 targeted reviews outside numbered rounds: 81 `triage_reviewed` and 2 `human-reviewed` rows. |
+| `sourceProposalID`, `sourcePoolIndex` | Identifiers linking a retained occurrence to its original proposal and embedding-pool record. |
+
+## `data/events.csv`
+
+Darwin Core Event hierarchy for the camera-trap component: one camera project,
+12 SensorDeployments, and 19,772 Image Events.
+
+## `data/images.csv`
+
+The full corrected inventory of 235,778 processed camera images, including
+images with no retained operational occurrences. `occurrence_count` counts
+rows in `data/occurrences.csv` associated with that image; zero does not
+establish animal absence. `raw_timestamp` preserves camera metadata, whereas
+`corrected_timestamp` is the timestamp used for Event and Occurrence analyses.
+Preview assets may retain embedded camera metadata; use `corrected_timestamp`
+for temporal analysis.
+
+Privacy fields preserve screening evidence and human-review outcome. The 31
+screening-flagged curated images with retained operational occurrences were
+human-reviewed and approved for upload.
+
+## `media_manifest.csv`
+
+One row per discovered media or auxiliary file. This manifest is broader than
+`data/images.csv`: it also includes videos and supporting files. `release_path`
+is a release-relative path; `source_path` is an internal acquisition path.
+`include = true` denotes intended package inclusion only and is **not** public
+release, licensing, or privacy-clearance authorization. Privacy-screening and
+human-review fields record the available evidence for each manifest entry.
+
+## `analysis/`
+
+`camera_image_acquisition_arrivals.csv` describes full-image acquisition.
+`camera_operational_occurrence_arrivals.csv` and
+`camera_arrival_statistics.json` describe the retained operational-occurrence
+workload.

smartwilds-camera-traps/metadata/camera_geolocation_config.csv

@@ -0,0 +1,14 @@
+enabled,collection,camera_id,path_contains,camera_lat,camera_lon,hfov_deg,bearing_deg,image_rotation_deg,target_height_m,config_source,notes
+1,thewilds_20250630_20250703,TW01-CT02,,39.83076500,-81.73789500,56.5,135,0,0.9,metadata_dms_plus_sensor_config,"GPS from metadata 39.49.8459 081.44.2737; HFOV/bearing from sensor_configurations.md"
+1,thewilds_20250630_20250703,TW02-CT03,,39.83065667,-81.73791833,63,225,0,0.9,metadata_dms_plus_sensor_config,"GPS from metadata 39.49.8394 081.44.2751; HFOV/bearing from sensor_configurations.md"
+1,thewilds_20250630_20250703,TW03-CT01,,39.83516667,-81.73518333,56.5,180,0,0.9,camera_log_dms_like_plus_sensor_config,"GPS normalized from camera log 39.5011 81.4411 as degrees+decimal-minutes; bearing/model from sensor_configurations.md"
+1,thewilds_20250630_20250703,TW04-CT04,,39.83438000,-81.73932000,120,315,180,0.9,camera_log_decimal_plus_sensor_config,"GPS from camera log; longitude sign corrected west; June/July camera facing NW from sensor configuration; images are upside down so bbox coordinates are mirrored during geolocation"
+1,thewilds_20250804_20250828,TW01-CT02,,39.83076500,-81.73789500,56.5,135,0,0.9,metadata_dms_plus_sensor_config,"GPS from metadata 39.49.8459 081.44.2737; HFOV/bearing from sensor_configurations.md"
+1,thewilds_20250804_20250828,TW02-CT01,,39.83065667,-81.73791833,56.5,225,0,0.9,metadata_dms_plus_sensor_config,"GPS from metadata 39.49.8394 081.44.2751; HFOV/bearing from sensor_configurations.md"
+1,thewilds_20250804_20250828,TW04-CT03,,39.83428667,-81.73934000,63,225,0,0.9,metadata_dms_plus_visual_review,"GPS from metadata 39.50.0572 081.44.3604; bearing corrected to southwest from visual review"
+1,thewilds_20250804_20250828,TW06-CT04,,39.83553983,-81.72999900,120,315,0,0.9,zebra_shelter_drone_frame_anchor,"Approximate Zebra Shelter camera anchor from 20250818_parrot_02 frame P5770586_000000 telemetry; HFOV/bearing from sensor_configurations.md"
+1,thewilds_20250804_20250828,TW06-CT05,,39.83553983,-81.72999900,55,0,0,0.9,zebra_shelter_drone_frame_anchor,"Approximate Zebra Shelter camera anchor from 20250818_parrot_02 frame P5770586_000000 telemetry; HFOV/bearing from sensor_configurations.md"
+1,thewilds_20250804_20250828,TW09-CT05,,39.83418000,-81.73783000,55,135,0,0.9,sensor_config_aru_site_proxy,"No direct camera GPS found; using TW09 Dock ARU/site coordinate; HFOV/bearing from sensor_configurations.md"
+1,thewilds_20250804_20250828,TW09_CT02,SD23_20250818_20250828,39.83065667,-81.73791833,56.5,225,0,0.9,relocated_to_nomad_ridge_proxy_needs_review,"Specific override for SD23 after move to Nomad Ridge/Shelter; GPS from TW02 metadata; bearing assumed from TW02/TW09 move notes"
+1,thewilds_20250804_20250828,TW09_CT02,,39.83418000,-81.73783000,56.5,180,0,0.9,sensor_config_aru_site_proxy,"Default TW09 Dock-left position; GPS from TW09 ARU/site coordinate; HFOV/bearing from sensor_configurations.md"
+1,thewilds_20250804_20250828,TW11-CT06,,39.83438000,-81.73930000,55,45,0,0.9,metadata_decimal_plus_sensor_config,"GPS from camera metadata 39.83438 -81.73930; HFOV/bearing from sensor_configurations.md"

smartwilds-camera-traps/metadata/camera_models.csv

@@ -0,0 +1,5 @@
+Model,Manufacturer,HFOV (°),Notes
+GardePro T5NG,GardePro,56.5,"Standard trail cam, photo-only or hybrid"
+GardePro A3,GardePro,63,Video-capable model
+Trail Camera G600,Trail Camera,120,Wide-angle lens
+Reveal Pro 3.0,Tactacam,55,"Cellular-capable, 4K photo, HD video"

smartwilds-camera-traps/metadata/deployments.csv

@@ -0,0 +1,15 @@
+deploymentID,Camera,Site,Model,HFOV,Facing,Mode,Photo Config,Video Config,PIR,Delay,Notes,campaign,PIR/Sensitivity,path_contains,camera_lat,camera_lon,hfov_deg,bearing_deg,image_rotation_deg,target_height_m,config_source,geolocation_notes,images_on_disk,released
+TW01-CT02,CT02,TW01 - Nomad Ridge E,GardePro T5NG,56.5°,SE (135°),Hybrid,"4MP, 3-burst","1080p, 10s, MP4+audio",High,—,"Timelapse hourly, info strip ON",thewilds_20250630_20250703,,,39.83076500,-81.73789500,56.5,135,0,0.9,metadata_dms_plus_sensor_config,GPS from metadata 39.49.8459 081.44.2737; HFOV/bearing from sensor_configurations.md,7811,true
+TW02-CT03,CT03,TW02 - Nomad Ridge W,GardePro A3,63°,SW (225°),Photo+Video,"8MP, 3-series","1080p, 10s, AVI+audio",High,10s shot lag,Only 55 files captured (early battery death),thewilds_20250630_20250703,,,39.83065667,-81.73791833,63,225,0,0.9,metadata_dms_plus_sensor_config,GPS from metadata 39.49.8394 081.44.2751; HFOV/bearing from sensor_configurations.md,39,true
+TW03-CT01,CT01,TW03 - Pasture D Feeder,GardePro T5NG,56.5°,S (180°),—,—,—,—,—,"Near salt lick, config not documented",thewilds_20250630_20250703,,,39.83516667,-81.73518333,56.5,180,0,0.9,camera_log_dms_like_plus_sensor_config,GPS normalized from camera log 39.5011 81.4411 as degrees+decimal-minutes; bearing/model from sensor_configurations.md,6661,true
+TW04-CT04,CT04,TW04 - Lake Trail,Trail Camera G600,120°,NW (315°),Photo+Video,"8MP, 5-burst","1080p, 10s, MP4+audio",High,—,Images upside-down (mount orientation),thewilds_20250630_20250703,,,39.83438000,-81.73932000,120,315,180,0.9,camera_log_decimal_plus_sensor_config,GPS from camera log; longitude sign corrected west; June/July camera facing NW from sensor configuration; images are upside down so bbox coordinates are mirrored during geolocation,570,true
+TW01-CT02,CT02,TW01 - Nomad Ridge E,GardePro T5NG,56.5°,SE (135°),Photo only,"4MP, 3-burst",—,,—,"26,573 JPG; removed Aug 11, relocated to dock",thewilds_20250804_20250828,High,,39.83076500,-81.73789500,56.5,135,0,0.9,metadata_dms_plus_sensor_config,GPS from metadata 39.49.8459 081.44.2737; HFOV/bearing from sensor_configurations.md,26573,true
+TW02-CT01,CT01,TW02 - Nomad Ridge W,GardePro T5NG,56.5°,SW (225°),Photo only,"4MP, 3-burst",—,,—,"61,217 JPG; DATE ERROR (clock incorrect); battery died Aug 18-25",thewilds_20250804_20250828,High,,39.83065667,-81.73791833,56.5,225,0,0.9,metadata_dms_plus_sensor_config,GPS from metadata 39.49.8394 081.44.2751; HFOV/bearing from sensor_configurations.md,61217,true
+TW04-CT03,CT03,TW04 - Lake Trail,GardePro A3,63°,NW (315°),Photo+Video,"8MP, 3-burst","1080p, 10s",,—,"984 JPG + 67 AVI; reset Aug 7, no data after Aug 25",thewilds_20250804_20250828,—,,39.83428667,-81.73934000,63,225,0,0.9,metadata_dms_plus_visual_review,GPS from metadata 39.50.0572 081.44.3604; bearing corrected to southwest from visual review,984,true
+TW06-CT04,CT04,TW06 - Zebra Shelter,Trail Camera G600,120°,NW (315°),Photo+Video,"8MP, 5-burst","1080p, 10s",,—,"40,907 JPG; battery 100% (unusual)",thewilds_20250804_20250828,Motion detection,,39.83553983,-81.72999900,120,315,0,0.9,zebra_shelter_drone_frame_anchor,Approximate Zebra Shelter camera anchor from 20250818_parrot_02 frame P5770586_000000 telemetry; HFOV/bearing from sensor_configurations.md,40907,true
+TW06-CT05,CT05,TW06 - Zebra Shelter (inside),Reveal Pro 3.0,55°,N (0°),Photo+Video,4K,"720p, 10s",,—,84 jpg + 84 mp4; late deploy Aug 25-28,thewilds_20250804_20250828,—,,39.83553983,-81.72999900,55,0,0,0.9,zebra_shelter_drone_frame_anchor,Approximate Zebra Shelter camera anchor from 20250818_parrot_02 frame P5770586_000000 telemetry; HFOV/bearing from sensor_configurations.md,84,true
+TW07-CT05,CT05,TW07 - Zipline Tower,Reveal Pro 3.0,55°,N (0°),Photo+Video,4K,"720p, 10s",,—,ALL PEOPLE (tour buses); no wildlife images,thewilds_20250804_20250828,—,,,,,,,,,,0,false
+TW09-CT02,CT02,TW09 - Dock (left),GardePro T5NG,56.5°,S (180°),Photo only,"4MP, 3-burst",—,,—,"90,834 JPG (largest dataset); battery gap Aug 18-25",thewilds_20250804_20250828,—,SD23_20250818_20250828,39.83065667,-81.73791833,56.5,225,0,0.9,relocated_to_nomad_ridge_proxy_needs_review,Specific override for SD23 after move to Nomad Ridge/Shelter; GPS from TW02 metadata; bearing assumed from TW02/TW09 move notes,59568,true
+TW09-CT02,CT02,TW09 - Dock (left),GardePro T5NG,56.5°,S (180°),Photo only,"4MP, 3-burst",—,,—,"90,834 JPG (largest dataset); battery gap Aug 18-25",thewilds_20250804_20250828,—,,39.83418000,-81.73783000,56.5,180,0,0.9,sensor_config_aru_site_proxy,Default TW09 Dock-left position; GPS from TW09 ARU/site coordinate; HFOV/bearing from sensor_configurations.md,31266,true
+TW09-CT05,CT05,TW09 - Dock (right),Reveal Pro 3.0,55°,SE (135°),Photo+Video,4K,"720p, 10s",,—,Only 9 jpg + 6 mp4; detection issues,thewilds_20250804_20250828,—,,39.83418000,-81.73783000,55,135,0,0.9,sensor_config_aru_site_proxy,No direct camera GPS found; using TW09 Dock ARU/site coordinate; HFOV/bearing from sensor_configurations.md,9,true
+TW11-CT06,CT06,TW11 - Fence/lake,Reveal Pro 3.0,55°,NE (45°),Photo+Video,4K,"FHD 1080p, 15s",,20s,"89 jpg + 23 mp4; Pic+Video mode, multi-shot 1s interval, flash high, 1 bar cell service",thewilds_20250804_20250828,Level 9,,39.83438000,-81.73930000,55,45,0,0.9,metadata_decimal_plus_sensor_config,GPS from camera metadata 39.83438 -81.73930; HFOV/bearing from sensor_configurations.md,89,true

smartwilds-camera-traps/metadata/provenance/ANNOTATION_PROVENANCE.md

@@ -0,0 +1,36 @@
+# Camera-trap annotation provenance
+
+Operational table: 72,775 retained animal boxes with original SAM3 geometry.
+
+## Label provenance
+
+| `labelProvenance` | Rows | Meaning |
+|---|---:|---|
+| `triage_resolved` | 70,728 | Machine label assigned by Resolve–Clear–Defer prototype triage. |
+| `triage_reviewed` | 1,807 | Human-reviewed triage label. Eighty-one rows have no numbered `triageRound`; these are targeted reviews outside the numbered rounds, including banteng. |
+| `triage_forced_1nn` | 199 | Small proposal unresolved after guarded Resolve–Clear–Defer triage; assigned its nearest reviewed-prototype label without the usual decision safeguards. It is a machine assignment and was not individually expert-verified. |
+| `human-edit` | 39 | Targeted human taxonomic correction to an existing SAM3 proposal. |
+| `human-reviewed` | 2 | Targeted human review label on an existing SAM3 proposal. |
+| **Total** | **72,775** | **Retained operational occurrences.** |
+
+## `triageRound`
+
+A non-empty `triageRound` identifies the numbered Resolve–Clear–Defer review
+round associated with that proposal. Eighty-three retained rows have no
+numbered round: 81 `triage_reviewed` targeted reviews and 2 `human-reviewed`
+records. Missing values therefore indicate review outside the numbered-round
+workflow, not missing annotation provenance.
+
+## Geometry provenance
+
+- `SAM3`: 72,775
+
+Every retained box preserves original SAM3 proposal geometry. Targeted human
+review or taxonomic correction changes label provenance only; it does not add
+or edit bounding-box geometry.
+
+## Verification scope
+
+The `triage_reviewed` rows are the human-reviewed triage subset. The remaining
+operational labels are semi-automated and should not be treated as a fully
+human-verified object-detection benchmark.

smartwilds-camera-traps/metadata/provenance/tw07_ct05_sensitive_media_disposition_20261008.json

@@ -0,0 +1,9 @@
+{
+  "deployment": "TW07-CT05",
+  "source_metadata": "thewilds_20250804_20250828/camera_traps/TW07-CT05/metadata.txt",
+  "retained_raw_media_files": 0,
+  "release_event_rows": 0,
+  "release_occurrence_rows": 0,
+  "release_decision": "Excluded people-only media because source metadata states photos were not uploaded.",
+  "not_claimed": "Whether any camera-card images were later deleted."
+}

smartwilds-camera-traps/metadata/taxonomy_mapping.csv

@@ -0,0 +1,14 @@
+sourceLabel,scientificName,vernacularName,taxonRank,phylum,class,order,family,genus
+banteng,Bos javanicus,banteng,species,Chordata,Mammalia,Artiodactyla,Bovidae,Bos
+barn swallow,Hirundo rustica,barn swallow,species,Chordata,Aves,Passeriformes,Hirundinidae,Hirundo
+bird,Aves,bird,class,Chordata,Aves,,,
+butterfly,Lepidoptera,butterfly,order,Arthropoda,Insecta,Lepidoptera,,
+deer,Elaphurus davidianus,milu deer,species,Chordata,Mammalia,Artiodactyla,Cervidae,Elaphurus
+dragonfly,Anisoptera,dragonfly,infraorder,Arthropoda,Insecta,Odonata,,
+e_s,Sturnus vulgaris,European starling,species,Chordata,Aves,Passeriformes,Sturnidae,Sturnus
+goose,Branta canadensis,Canada goose,species,Chordata,Aves,Anseriformes,Anatidae,Branta
+horse,Equus przewalskii,Przewalski's horse,species,Chordata,Mammalia,Perissodactyla,Equidae,Equus
+insect,Insecta,insect,class,Arthropoda,Insecta,,,
+killdeer,Charadrius vociferus,killdeer,species,Chordata,Aves,Charadriiformes,Charadriidae,Charadrius
+swan,Cygnus buccinator,trumpeter swan,species,Chordata,Aves,Anseriformes,Anatidae,Cygnus
+takin,Budorcas taxicolor,Sichuan takin,species,Chordata,Mammalia,Artiodactyla,Bovidae,Budorcas

smartwilds-camera-traps/metadata/timestamp_summary_by_camera.csv

@@ -0,0 +1,13 @@
+dataset,camera_id,location,images,mapped_images,unmapped_images,first_corrected_timestamp,last_corrected_timestamp,needs_review_images,raw_source_counts,correction_method_counts
+thewilds_20250630_20250703,TW01-CT02,Nomad Ridge E,7811,7811,0,2025-06-30T12:18:02,2025-07-02T06:50:02,0,"{""filename"": 7811}","{""june_tw01_correct"": 7811}"
+thewilds_20250630_20250703,TW02-CT03,Nomad Ridge W,39,39,0,2025-06-30T12:34:39,2025-07-01T00:34:44,0,"{""exif"": 39}","{""june_tw02_start_anchor"": 39}"
+thewilds_20250630_20250703,TW03-CT01,Pasture D Feeder,6661,6661,0,2025-06-30T13:06:04,2025-07-02T01:04:01,0,"{""filename"": 6661}","{""june_tw03_correct"": 6661}"
+thewilds_20250630_20250703,TW04-CT04,Lake Trail,570,570,0,2025-06-30T13:25:24,2025-07-03T16:56:10,0,"{""exif"": 570}","{""june_tw04_start_anchor_inverted_images"": 570}"
+thewilds_20250804_20250828,TW01-CT02,Nomad Ridge E,26573,26573,0,2025-08-04T09:01:26,2025-08-10T10:15:50,0,"{""filename"": 26573}","{""aug_tw01_2023_tail_clock_offset"": 15, ""aug_tw01_2025_clock_offset"": 26558}"
+thewilds_20250804_20250828,TW02-CT01,Nomad Ridge W,61217,61217,0,2025-08-04T09:05:00,2025-08-18T09:07:53,0,"{""filename"": 61217}","{""aug_tw02_actual_august_clock"": 17301, ""aug_tw02_sd01_2023_clock_anchor"": 13689, ""aug_tw02_sd01_sep_clock_anchor"": 17262, ""aug_tw02_sd11_sep_clock_anchor"": 12965}"
+thewilds_20250804_20250828,TW04-CT03,Lake Trail,984,984,0,2025-08-07T10:38:38,2025-08-25T10:49:44,0,"{""exif"": 984}","{""aug_tw04_sd02_start_anchor"": 696, ""aug_tw04_sd03_start_anchor"": 288}"
+thewilds_20250804_20250828,TW06-CT04,Zebra Shelter,40907,40907,0,2025-08-04T09:52:35,2025-08-26T15:35:09,0,"{""exif"": 40907}","{""aug_tw06_ct04_initial_start_anchor"": 13173, ""aug_tw06_ct04_sd02_start_anchor"": 27734}"
+thewilds_20250804_20250828,TW06-CT05,Zebra Shelter,84,84,0,2025-08-25T15:09:52,2025-08-28T08:17:12,0,"{""bottom_overlay_ocr"": 83, ""manual_bottom_overlay_read"": 1}","{""bottom_overlay_ocr"": 83, ""manual_bottom_overlay_read"": 1}"
+thewilds_20250804_20250828,TW09-CT05,Dock / Waterfront,9,9,0,2025-08-11T10:18:09,2025-08-13T18:20:26,0,"{""bottom_overlay_ocr"": 9}","{""bottom_overlay_ocr"": 9}"
+thewilds_20250804_20250828,TW09_CT02,"Dock / Waterfront, then Nomad Ridge W after 2025-08-18",90834,90834,0,2025-08-11T10:27:33,2025-08-28T09:44:43,0,"{""filename"": 90834}","{""aug_tw09_ct02_sd22_start_anchor"": 31266, ""aug_tw09_ct02_sd23_correct_after_move"": 59568}"
+thewilds_20250804_20250828,TW11-CT06,Nomad Ridge W,89,89,0,2025-08-18T09:00:37,2025-08-28T02:47:24,0,"{""bottom_overlay_ocr"": 87, ""manual_bottom_overlay_read"": 2}","{""bottom_overlay_ocr"": 87, ""manual_bottom_overlay_read"": 2}"

smartwilds-camera-traps/metadata/timestamp_summary_by_date.csv

@@ -0,0 +1,117 @@
+dataset,camera_id,sd_folder,corrected_date,correction_method,images,first_corrected_timestamp,last_corrected_timestamp,media_folders
+thewilds_20250630_20250703,TW01-CT02,SD02_20250630_20250703,2025-06-30,june_tw01_correct,2142,2025-06-30T12:18:02,2025-06-30T23:59:02,chunk1;chunk2;chunk3;chunk4
+thewilds_20250630_20250703,TW01-CT02,SD02_20250630_20250703,2025-07-01,june_tw01_correct,4392,2025-07-01T00:00:01,2025-07-01T23:59:02,chunk1;chunk2;chunk3;chunk4
+thewilds_20250630_20250703,TW01-CT02,SD02_20250630_20250703,2025-07-02,june_tw01_correct,1277,2025-07-02T00:00:02,2025-07-02T06:50:02,chunk1;chunk2;chunk3;chunk4
+thewilds_20250630_20250703,TW02-CT03,SD03_20250630_20250703,2025-06-30,june_tw02_start_anchor,36,2025-06-30T12:34:39,2025-06-30T23:34:47,images
+thewilds_20250630_20250703,TW02-CT03,SD03_20250630_20250703,2025-07-01,june_tw02_start_anchor,3,2025-07-01T00:34:43,2025-07-01T00:34:44,images
+thewilds_20250630_20250703,TW03-CT01,SD01_20250630_20250703,2025-06-30,june_tw03_correct,1979,2025-06-30T13:06:04,2025-06-30T23:59:01,100MEDIA;101MEDIA;102MEDIA
+thewilds_20250630_20250703,TW03-CT01,SD01_20250630_20250703,2025-07-01,june_tw03_correct,4487,2025-07-01T00:00:01,2025-07-01T23:59:01,102MEDIA;103MEDIA;104MEDIA;105MEDIA;106MEDIA;107MEDIA;108MEDIA
+thewilds_20250630_20250703,TW03-CT01,SD01_20250630_20250703,2025-07-02,june_tw03_correct,195,2025-07-02T00:00:01,2025-07-02T01:04:01,108MEDIA
+thewilds_20250630_20250703,TW04-CT04,SD04_20250630_20250704,2025-06-30,june_tw04_start_anchor_inverted_images,40,2025-06-30T13:25:24,2025-06-30T18:18:31,101MEDIA
+thewilds_20250630_20250703,TW04-CT04,SD04_20250630_20250704,2025-07-01,june_tw04_start_anchor_inverted_images,250,2025-07-01T18:42:33,2025-07-01T20:16:41,101MEDIA
+thewilds_20250630_20250703,TW04-CT04,SD04_20250630_20250704,2025-07-02,june_tw04_start_anchor_inverted_images,255,2025-07-02T07:48:38,2025-07-02T20:17:58,101MEDIA
+thewilds_20250630_20250703,TW04-CT04,SD04_20250630_20250704,2025-07-03,june_tw04_start_anchor_inverted_images,25,2025-07-03T15:36:22,2025-07-03T16:56:10,101MEDIA
+thewilds_20250804_20250828,TW01-CT02,SD02_20250804_20250811,2025-08-04,aug_tw01_2025_clock_offset,999,2025-08-04T09:01:26,2025-08-04T14:33:27,109MEDIA
+thewilds_20250804_20250828,TW01-CT02,SD02_20250804_20250811,2025-08-05,aug_tw01_2025_clock_offset,4416,2025-08-05T01:40:26,2025-08-05T23:59:27,112MEDIA;113MEDIA;114MEDIA;115MEDIA;116MEDIA
+thewilds_20250804_20250828,TW01-CT02,SD02_20250804_20250811,2025-08-06,aug_tw01_2025_clock_offset,4437,2025-08-06T00:00:26,2025-08-06T23:59:27,116MEDIA;117MEDIA;118MEDIA;119MEDIA;120MEDIA
+thewilds_20250804_20250828,TW01-CT02,SD02_20250804_20250811,2025-08-07,aug_tw01_2025_clock_offset,4764,2025-08-07T00:00:26,2025-08-07T23:59:27,120MEDIA;121MEDIA;122MEDIA;123MEDIA;124MEDIA;125MEDIA
+thewilds_20250804_20250828,TW01-CT02,SD02_20250804_20250811,2025-08-08,aug_tw01_2025_clock_offset,4680,2025-08-08T00:00:26,2025-08-08T23:59:27,125MEDIA;126MEDIA;127MEDIA;128MEDIA;129MEDIA;130MEDIA
+thewilds_20250804_20250828,TW01-CT02,SD02_20250804_20250811,2025-08-09,aug_tw01_2025_clock_offset,4910,2025-08-09T00:00:26,2025-08-09T23:59:27,130MEDIA;131MEDIA;132MEDIA;133MEDIA;134MEDIA;135MEDIA
+thewilds_20250804_20250828,TW01-CT02,SD02_20250804_20250811,2025-08-10,aug_tw01_2023_tail_clock_offset,15,2025-08-10T10:14:41,2025-08-10T10:15:50,137MEDIA
+thewilds_20250804_20250828,TW01-CT02,SD02_20250804_20250811,2025-08-10,aug_tw01_2025_clock_offset,2352,2025-08-10T00:00:26,2025-08-10T10:14:31,135MEDIA;136MEDIA;137MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD01_20250804_20250811,2025-08-04,aug_tw02_sd01_2023_clock_anchor,8388,2025-08-04T09:05:00,2025-08-04T23:59:01,100MEDIA;101MEDIA;102MEDIA;104MEDIA;105MEDIA;106MEDIA;107MEDIA;108MEDIA;109MEDIA;110MEDIA;111MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD01_20250804_20250811,2025-08-05,aug_tw02_sd01_2023_clock_anchor,5301,2025-08-05T00:00:00,2025-08-05T09:04:01,100MEDIA;102MEDIA;103MEDIA;104MEDIA;107MEDIA;108MEDIA;111MEDIA;112MEDIA;113MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD01_20250804_20250811,2025-08-07,aug_tw02_sd01_sep_clock_anchor,2676,2025-08-07T09:09:02,2025-08-07T23:59:01,113MEDIA;114MEDIA;115MEDIA;116MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD01_20250804_20250811,2025-08-08,aug_tw02_sd01_sep_clock_anchor,4320,2025-08-08T00:00:01,2025-08-08T23:59:02,116MEDIA;117MEDIA;118MEDIA;119MEDIA;120MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD01_20250804_20250811,2025-08-09,aug_tw02_sd01_sep_clock_anchor,4338,2025-08-09T00:00:01,2025-08-09T23:59:02,120MEDIA;121MEDIA;122MEDIA;123MEDIA;124MEDIA;125MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD01_20250804_20250811,2025-08-10,aug_tw02_sd01_sep_clock_anchor,4320,2025-08-10T00:00:01,2025-08-10T23:59:02,125MEDIA;126MEDIA;127MEDIA;128MEDIA;129MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD01_20250804_20250811,2025-08-11,aug_tw02_sd01_sep_clock_anchor,1608,2025-08-11T00:00:01,2025-08-11T08:50:34,129MEDIA;130MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD11_20250811_20250818,2025-08-11,aug_tw02_sd11_sep_clock_anchor,2624,2025-08-11T09:34:00,2025-08-11T23:59:34,100MEDIA;101MEDIA;102MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD11_20250811_20250818,2025-08-12,aug_tw02_sd11_sep_clock_anchor,4329,2025-08-12T00:00:33,2025-08-12T23:59:34,102MEDIA;103MEDIA;104MEDIA;105MEDIA;106MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD11_20250811_20250818,2025-08-13,aug_tw02_sd11_sep_clock_anchor,4344,2025-08-13T00:00:33,2025-08-13T23:59:34,106MEDIA;107MEDIA;108MEDIA;109MEDIA;110MEDIA;111MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD11_20250811_20250818,2025-08-14,aug_tw02_actual_august_clock,2646,2025-08-14T09:18:22,2025-08-14T23:59:01,112MEDIA;113MEDIA;114MEDIA;115MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD11_20250811_20250818,2025-08-14,aug_tw02_sd11_sep_clock_anchor,1668,2025-08-14T00:00:33,2025-08-14T09:12:44,111MEDIA;112MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD11_20250811_20250818,2025-08-15,aug_tw02_actual_august_clock,4353,2025-08-15T00:00:00,2025-08-15T23:59:01,115MEDIA;116MEDIA;117MEDIA;118MEDIA;119MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD11_20250811_20250818,2025-08-16,aug_tw02_actual_august_clock,4323,2025-08-16T00:00:00,2025-08-16T23:59:01,119MEDIA;120MEDIA;121MEDIA;122MEDIA;123MEDIA;124MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD11_20250811_20250818,2025-08-17,aug_tw02_actual_august_clock,4332,2025-08-17T00:00:00,2025-08-17T23:59:01,124MEDIA;125MEDIA;126MEDIA;127MEDIA;128MEDIA
+thewilds_20250804_20250828,TW02-CT01,SD11_20250811_20250818,2025-08-18,aug_tw02_actual_august_clock,1647,2025-08-18T00:00:00,2025-08-18T09:07:53,128MEDIA;129MEDIA;130MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD02_20250811_20250828,2025-08-11,aug_tw04_sd02_start_anchor,42,2025-08-11T10:08:18,2025-08-11T23:08:23,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD02_20250811_20250828,2025-08-12,aug_tw04_sd02_start_anchor,72,2025-08-12T00:08:23,2025-08-12T23:08:44,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD02_20250811_20250828,2025-08-13,aug_tw04_sd02_start_anchor,54,2025-08-13T00:08:44,2025-08-13T20:09:02,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD02_20250811_20250828,2025-08-14,aug_tw04_sd02_start_anchor,12,2025-08-14T07:09:10,2025-08-14T10:09:13,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD02_20250811_20250828,2025-08-18,aug_tw04_sd02_start_anchor,39,2025-08-18T11:18:44,2025-08-18T23:18:56,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD02_20250811_20250828,2025-08-19,aug_tw04_sd02_start_anchor,72,2025-08-19T00:18:55,2025-08-19T23:19:15,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD02_20250811_20250828,2025-08-20,aug_tw04_sd02_start_anchor,72,2025-08-20T00:19:14,2025-08-20T23:19:37,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD02_20250811_20250828,2025-08-21,aug_tw04_sd02_start_anchor,72,2025-08-21T00:19:36,2025-08-21T23:19:55,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD02_20250811_20250828,2025-08-22,aug_tw04_sd02_start_anchor,72,2025-08-22T00:19:53,2025-08-22T23:20:08,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD02_20250811_20250828,2025-08-23,aug_tw04_sd02_start_anchor,72,2025-08-23T00:20:07,2025-08-23T23:20:32,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD02_20250811_20250828,2025-08-24,aug_tw04_sd02_start_anchor,72,2025-08-24T00:20:32,2025-08-24T23:21:01,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD02_20250811_20250828,2025-08-25,aug_tw04_sd02_start_anchor,45,2025-08-25T00:21:00,2025-08-25T10:49:44,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD03_20250807_20250811,2025-08-07,aug_tw04_sd03_start_anchor,42,2025-08-07T10:38:38,2025-08-07T23:38:44,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD03_20250807_20250811,2025-08-08,aug_tw04_sd03_start_anchor,72,2025-08-08T00:38:44,2025-08-08T23:39:14,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD03_20250807_20250811,2025-08-09,aug_tw04_sd03_start_anchor,72,2025-08-09T00:39:14,2025-08-09T23:39:42,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD03_20250807_20250811,2025-08-10,aug_tw04_sd03_start_anchor,72,2025-08-10T00:39:42,2025-08-10T23:40:00,100MEDIA
+thewilds_20250804_20250828,TW04-CT03,SD03_20250807_20250811,2025-08-11,aug_tw04_sd03_start_anchor,30,2025-08-11T00:40:00,2025-08-11T09:40:14,100MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-11,aug_tw06_ct04_sd02_start_anchor,1890,2025-08-11T09:48:51,2025-08-11T21:16:28,100MEDIA;101MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-12,aug_tw06_ct04_sd02_start_anchor,2170,2025-08-12T01:51:41,2025-08-12T18:21:45,101MEDIA;102MEDIA;103MEDIA;104MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-13,aug_tw06_ct04_sd02_start_anchor,520,2025-08-13T03:15:44,2025-08-13T21:58:11,104MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-14,aug_tw06_ct04_sd02_start_anchor,972,2025-08-14T01:31:12,2025-08-14T23:37:33,104MEDIA;105MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-15,aug_tw06_ct04_sd02_start_anchor,875,2025-08-15T08:40:29,2025-08-15T23:39:55,105MEDIA;106MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-16,aug_tw06_ct04_sd02_start_anchor,670,2025-08-16T00:49:21,2025-08-16T17:35:32,106MEDIA;107MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-17,aug_tw06_ct04_sd02_start_anchor,2890,2025-08-17T01:15:17,2025-08-17T22:43:08,107MEDIA;108MEDIA;109MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-18,aug_tw06_ct04_sd02_start_anchor,1935,2025-08-18T02:03:42,2025-08-18T20:31:33,109MEDIA;110MEDIA;111MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-19,aug_tw06_ct04_sd02_start_anchor,2015,2025-08-19T05:38:44,2025-08-19T21:48:46,111MEDIA;112MEDIA;113MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-20,aug_tw06_ct04_sd02_start_anchor,695,2025-08-20T02:38:58,2025-08-20T23:29:53,113MEDIA;114MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-21,aug_tw06_ct04_sd02_start_anchor,435,2025-08-21T01:43:13,2025-08-21T18:55:12,114MEDIA;115MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-22,aug_tw06_ct04_sd02_start_anchor,1450,2025-08-22T04:33:23,2025-08-22T17:05:25,115MEDIA;116MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-23,aug_tw06_ct04_sd02_start_anchor,1740,2025-08-23T08:17:13,2025-08-23T23:59:48,116MEDIA;117MEDIA;118MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-24,aug_tw06_ct04_sd02_start_anchor,1780,2025-08-24T00:21:59,2025-08-24T18:37:05,118MEDIA;119MEDIA;120MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-25,aug_tw06_ct04_sd02_start_anchor,4235,2025-08-25T02:24:41,2025-08-25T21:44:36,120MEDIA;121MEDIA;122MEDIA;123MEDIA;124MEDIA
+thewilds_20250804_20250828,TW06-CT04,SD02_20250811_20250828,2025-08-26,aug_tw06_ct04_sd02_start_anchor,3462,2025-08-26T02:00:39,2025-08-26T15:35:09,124MEDIA;125MEDIA;126MEDIA;127MEDIA
+thewilds_20250804_20250828,TW06-CT04,images,2025-08-04,aug_tw06_ct04_initial_start_anchor,1990,2025-08-04T09:52:35,2025-08-04T18:21:47,100MEDIA;101MEDIA;102MEDIA
+thewilds_20250804_20250828,TW06-CT04,images,2025-08-05,aug_tw06_ct04_initial_start_anchor,1080,2025-08-05T03:55:47,2025-08-05T21:23:30,102MEDIA;103MEDIA
+thewilds_20250804_20250828,TW06-CT04,images,2025-08-06,aug_tw06_ct04_initial_start_anchor,3180,2025-08-06T04:19:18,2025-08-06T21:29:19,103MEDIA;104MEDIA;105MEDIA;106MEDIA
+thewilds_20250804_20250828,TW06-CT04,images,2025-08-07,aug_tw06_ct04_initial_start_anchor,2250,2025-08-07T08:33:44,2025-08-07T21:29:47,106MEDIA;107MEDIA;108MEDIA
+thewilds_20250804_20250828,TW06-CT04,images,2025-08-08,aug_tw06_ct04_initial_start_anchor,2865,2025-08-08T02:13:47,2025-08-08T18:00:35,108MEDIA;109MEDIA;110MEDIA;111MEDIA
+thewilds_20250804_20250828,TW06-CT04,images,2025-08-09,aug_tw06_ct04_initial_start_anchor,1620,2025-08-09T01:44:32,2025-08-09T23:12:14,111MEDIA;112MEDIA;113MEDIA
+thewilds_20250804_20250828,TW06-CT04,images,2025-08-10,aug_tw06_ct04_initial_start_anchor,188,2025-08-10T07:27:49,2025-08-10T11:16:58,113MEDIA
+thewilds_20250804_20250828,TW06-CT05,SD05_20250825_20250828,2025-08-25,bottom_overlay_ocr,16,2025-08-25T15:09:52,2025-08-25T18:56:02,images
+thewilds_20250804_20250828,TW06-CT05,SD05_20250825_20250828,2025-08-25,manual_bottom_overlay_read,1,2025-08-25T18:51:19,2025-08-25T18:51:19,images
+thewilds_20250804_20250828,TW06-CT05,SD05_20250825_20250828,2025-08-26,bottom_overlay_ocr,23,2025-08-26T01:09:24,2025-08-26T14:10:16,images
+thewilds_20250804_20250828,TW06-CT05,SD05_20250825_20250828,2025-08-27,bottom_overlay_ocr,17,2025-08-27T09:42:12,2025-08-27T21:52:40,images
+thewilds_20250804_20250828,TW06-CT05,SD05_20250825_20250828,2025-08-28,bottom_overlay_ocr,27,2025-08-28T00:20:24,2025-08-28T08:17:12,images
+thewilds_20250804_20250828,TW09-CT05,SD05_20250811_20250814,2025-08-11,bottom_overlay_ocr,2,2025-08-11T10:18:09,2025-08-11T22:18:41,SD05_20250811_20250814
+thewilds_20250804_20250828,TW09-CT05,SD05_20250811_20250814,2025-08-12,bottom_overlay_ocr,4,2025-08-12T16:35:32,2025-08-12T22:33:17,SD05_20250811_20250814
+thewilds_20250804_20250828,TW09-CT05,SD05_20250811_20250814,2025-08-13,bottom_overlay_ocr,3,2025-08-13T16:40:53,2025-08-13T18:20:26,SD05_20250811_20250814
+thewilds_20250804_20250828,TW09_CT02,SD22_20250811_20250818,2025-08-11,aug_tw09_ct02_sd22_start_anchor,2718,2025-08-11T10:27:33,2025-08-11T23:59:01,100MEDIA;101MEDIA;102MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD22_20250811_20250818,2025-08-12,aug_tw09_ct02_sd22_start_anchor,4617,2025-08-12T00:00:00,2025-08-12T23:59:01,102MEDIA;103MEDIA;104MEDIA;105MEDIA;106MEDIA;107MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD22_20250811_20250818,2025-08-13,aug_tw09_ct02_sd22_start_anchor,4353,2025-08-13T00:00:00,2025-08-13T23:59:01,107MEDIA;108MEDIA;109MEDIA;110MEDIA;111MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD22_20250811_20250818,2025-08-14,aug_tw09_ct02_sd22_start_anchor,4335,2025-08-14T00:00:00,2025-08-14T23:59:01,111MEDIA;112MEDIA;113MEDIA;114MEDIA;115MEDIA;116MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD22_20250811_20250818,2025-08-15,aug_tw09_ct02_sd22_start_anchor,4365,2025-08-15T00:00:00,2025-08-15T23:59:01,116MEDIA;117MEDIA;118MEDIA;119MEDIA;120MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD22_20250811_20250818,2025-08-16,aug_tw09_ct02_sd22_start_anchor,4524,2025-08-16T00:00:01,2025-08-16T23:59:01,120MEDIA;121MEDIA;122MEDIA;123MEDIA;124MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD22_20250811_20250818,2025-08-17,aug_tw09_ct02_sd22_start_anchor,4485,2025-08-17T00:00:00,2025-08-17T23:59:01,124MEDIA;125MEDIA;126MEDIA;127MEDIA;128MEDIA;129MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD22_20250811_20250818,2025-08-18,aug_tw09_ct02_sd22_start_anchor,1869,2025-08-18T00:00:00,2025-08-18T10:11:30,129MEDIA;130MEDIA;131MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD23_20250818_20250828,2025-08-18,aug_tw09_ct02_sd23_correct_after_move,2673,2025-08-18T09:17:01,2025-08-18T23:59:01,100MEDIA;101MEDIA;102MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD23_20250818_20250828,2025-08-19,aug_tw09_ct02_sd23_correct_after_move,4329,2025-08-19T00:00:00,2025-08-19T23:59:01,102MEDIA;103MEDIA;104MEDIA;105MEDIA;106MEDIA;107MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD23_20250818_20250828,2025-08-20,aug_tw09_ct02_sd23_correct_after_move,4341,2025-08-20T00:00:00,2025-08-20T23:59:01,107MEDIA;108MEDIA;109MEDIA;110MEDIA;111MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD23_20250818_20250828,2025-08-21,aug_tw09_ct02_sd23_correct_after_move,4332,2025-08-21T00:00:00,2025-08-21T23:59:01,111MEDIA;112MEDIA;113MEDIA;114MEDIA;115MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD23_20250818_20250828,2025-08-22,aug_tw09_ct02_sd23_correct_after_move,4455,2025-08-22T00:00:00,2025-08-22T23:59:01,115MEDIA;116MEDIA;117MEDIA;118MEDIA;119MEDIA;120MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD23_20250818_20250828,2025-08-23,aug_tw09_ct02_sd23_correct_after_move,4380,2025-08-23T00:00:00,2025-08-23T23:59:01,120MEDIA;121MEDIA;122MEDIA;123MEDIA;124MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD23_20250818_20250828,2025-08-24,aug_tw09_ct02_sd23_correct_after_move,4359,2025-08-24T00:00:01,2025-08-24T23:59:01,124MEDIA;125MEDIA;126MEDIA;127MEDIA;128MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD23_20250818_20250828,2025-08-25,aug_tw09_ct02_sd23_correct_after_move,7564,2025-08-25T00:00:00,2025-08-25T23:59:01,100MEDIA;101MEDIA;102MEDIA;103MEDIA;128MEDIA;129MEDIA;130MEDIA;131MEDIA;132MEDIA;133MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD23_20250818_20250828,2025-08-26,aug_tw09_ct02_sd23_correct_after_move,9429,2025-08-26T00:00:00,2025-08-26T23:59:01,103MEDIA;104MEDIA;105MEDIA;106MEDIA;107MEDIA;108MEDIA;133MEDIA;134MEDIA;135MEDIA;136MEDIA;137MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD23_20250818_20250828,2025-08-27,aug_tw09_ct02_sd23_correct_after_move,9468,2025-08-27T00:00:00,2025-08-27T23:59:01,108MEDIA;109MEDIA;110MEDIA;111MEDIA;112MEDIA;113MEDIA;137MEDIA;138MEDIA;139MEDIA;140MEDIA;141MEDIA;142MEDIA
+thewilds_20250804_20250828,TW09_CT02,SD23_20250818_20250828,2025-08-28,aug_tw09_ct02_sd23_correct_after_move,4238,2025-08-28T00:00:00,2025-08-28T09:44:43,113MEDIA;114MEDIA;115MEDIA;142MEDIA;143MEDIA
+thewilds_20250804_20250828,TW11-CT06,SD06_20250818_20250828,2025-08-18,bottom_overlay_ocr,8,2025-08-18T09:00:37,2025-08-18T21:09:47,images
+thewilds_20250804_20250828,TW11-CT06,SD06_20250818_20250828,2025-08-19,bottom_overlay_ocr,8,2025-08-19T09:09:30,2025-08-19T21:09:35,images
+thewilds_20250804_20250828,TW11-CT06,SD06_20250818_20250828,2025-08-20,bottom_overlay_ocr,2,2025-08-20T09:09:17,2025-08-20T21:09:18,images
+thewilds_20250804_20250828,TW11-CT06,SD06_20250818_20250828,2025-08-21,bottom_overlay_ocr,2,2025-08-21T09:08:58,2025-08-21T21:09:02,images
+thewilds_20250804_20250828,TW11-CT06,SD06_20250818_20250828,2025-08-22,bottom_overlay_ocr,8,2025-08-22T09:08:43,2025-08-22T21:08:46,images
+thewilds_20250804_20250828,TW11-CT06,SD06_20250818_20250828,2025-08-23,bottom_overlay_ocr,20,2025-08-23T08:57:32,2025-08-23T21:22:27,images
+thewilds_20250804_20250828,TW11-CT06,SD06_20250818_20250828,2025-08-24,bottom_overlay_ocr,2,2025-08-24T09:22:08,2025-08-24T21:22:13,images
+thewilds_20250804_20250828,TW11-CT06,SD06_20250818_20250828,2025-08-25,bottom_overlay_ocr,19,2025-08-25T03:58:41,2025-08-25T21:23:28,images
+thewilds_20250804_20250828,TW11-CT06,SD06_20250818_20250828,2025-08-25,manual_bottom_overlay_read,1,2025-08-25T09:00:36,2025-08-25T09:00:36,images
+thewilds_20250804_20250828,TW11-CT06,SD06_20250818_20250828,2025-08-26,bottom_overlay_ocr,2,2025-08-26T09:23:05,2025-08-26T21:23:09,images
+thewilds_20250804_20250828,TW11-CT06,SD06_20250818_20250828,2025-08-27,bottom_overlay_ocr,7,2025-08-27T08:01:01,2025-08-27T21:22:53,images
+thewilds_20250804_20250828,TW11-CT06,SD06_20250818_20250828,2025-08-27,manual_bottom_overlay_read,1,2025-08-27T09:23:47,2025-08-27T09:23:47,images
+thewilds_20250804_20250828,TW11-CT06,SD06_20250818_20250828,2025-08-28,bottom_overlay_ocr,9,2025-08-28T02:43:19,2025-08-28T02:47:24,images

smartwilds-camera-traps/scripts/generate_darwin_core.py

@@ -0,0 +1,1207 @@
+#!/usr/bin/env python3
+"""
+Generate the Darwin Core Event and Occurrence tables for The Wilds 2025.
+
+This script uses the existing modality-specific Darwin Core generators as the
+schema/style reference, but writes a fresh non-destructive set of tables under
+``SMARTWILDS_DWC_OUT`` (default ``<archive>/processing_output/darwin_core_release``).
+The field archive root is read from ``SMARTWILDS_ARCHIVE``; the external
+SAM3/BioCLIP detection workspace from ``SMARTWILDS_DETECTIONS``.
+
+Key updates relative to the earlier outputs:
+  - camera traps are regenerated from SAM3 + BioCLIP Tree-of-Life prediction
+    CSVs with geolocated animal-position estimates
+  - camera-trap timestamps use the current camera_image_timestamp_map.csv
+  - bioacoustic scientificName/genus/specificEpithet fields are filled from
+    raw BirdNET result files
+  - drone occurrences are regenerated from the labeled drone detection CSVs
+  - citizen science and text Darwin Core outputs are copied into the snapshot
+  - all modality-specific files are merged into events_all_modalities.csv and
+    occurrences_all_modalities.csv
+"""
+
+from __future__ import annotations
+
+import argparse
+import csv
+import json
+import math
+import os
+import re
+import shutil
+from collections import Counter, defaultdict
+from datetime import date, datetime
+from pathlib import Path
+
+
+BASE_DIR = Path(os.environ.get("SMARTWILDS_ARCHIVE", ".")).resolve()
+OUT_DIR = Path(os.environ.get("SMARTWILDS_DWC_OUT", BASE_DIR / "processing_output" / "darwin_core_release"))
+SMARTWILDS_DIR = Path(os.environ.get("SMARTWILDS_DETECTIONS", BASE_DIR / "detections_external"))
+
+EVENT_FIELDS = [
+    "eventID", "parentEventID", "eventType", "eventDate", "samplingProtocol",
+    "sampleSizeValue", "sampleSizeUnit", "locality", "locationID",
+    "decimalLatitude", "decimalLongitude", "coordinateUncertaintyInMeters",
+    "geodeticDatum", "country", "stateProvince", "county", "datasetID",
+    "datasetName", "recordedBy", "fieldNotes", "eventTime",
+]
+
+OCC_FIELDS = [
+    "occurrenceID", "eventID", "catalogNumber", "scientificName", "kingdom",
+    "phylum", "class", "order", "family", "genus", "specificEpithet",
+    "taxonRank", "vernacularName", "occurrenceStatus", "individualCount",
+    "lifeStage", "sex", "behavior", "basisOfRecord", "recordedBy",
+    "identifiedBy", "identificationVerificationStatus", "identificationRemarks",
+    "eventDate", "eventTime", "decimalLatitude", "decimalLongitude",
+    "coordinateUncertaintyInMeters", "geodeticDatum", "locality", "locationID",
+    "country", "stateProvince", "county", "establishmentMeans",
+    "degreeOfEstablishment", "occurrenceRemarks", "associatedMedia",
+    "associatedReferences", "datasetID", "datasetName", "informationWithheld",
+    "dataGeneralizations",
+]
+
+SITE_GPS = {
+    "TW01": ("39.83077", "-81.73789"),
+    "TW02": ("39.83066", "-81.73792"),
+    "TW03": ("39.83670", "-81.73607"),
+    "TW04": ("39.83429", "-81.73934"),
+    "TW05": ("39.83537", "-81.73842"),
+    "TW06": ("39.83554", "-81.73000"),
+    "TW07": ("39.83288", "-81.73455"),
+    "TW08": ("39.82900", "-81.74378"),
+    "TW09": ("39.82798", "-81.74167"),
+    "TW11": ("39.83438", "-81.73930"),
+}
+
+SITE_NAMES = {
+    "TW01": "Nomad Ridge Shelter (East)",
+    "TW02": "Nomad Ridge Shelter (West)",
+    "TW03": "Old Giraffe Feeder, Pasture D",
+    "TW04": "Lake Trail",
+    "TW05": "Lake Trail Gate O",
+    "TW06": "Zebra Shelter",
+    "TW07": "Zipline Tower",
+    "TW08": "Fence with Vegetation",
+    "TW09": "Dock",
+    "TW11": "Fence overlooking Lake",
+}
+
+STUDY_ROOT_EVENT = {
+    "eventID": "TheWilds_2025",
+    "parentEventID": "",
+    "eventType": "Project",
+    "eventDate": "2025-06-30/2025-09-08",
+    "samplingProtocol": "Multi-modal biodiversity survey across camera traps, bioacoustics, drone video, citizen science, and text observations.",
+    "sampleSizeValue": "5",
+    "sampleSizeUnit": "modality projects",
+    "locality": "The Wilds Conservation Center, Muskingum County, Ohio, USA",
+    "locationID": "",
+    "decimalLatitude": "39.83",
+    "decimalLongitude": "-81.74",
+    "coordinateUncertaintyInMeters": "5000",
+    "geodeticDatum": "WGS84",
+    "country": "United States",
+    "stateProvince": "Ohio",
+    "county": "Muskingum",
+    "datasetID": "TheWilds_2025",
+    "datasetName": "The Wilds 2025 Multi-Modal Survey",
+    "recordedBy": "Field team",
+    "fieldNotes": "Study-level root event for all modality project events.",
+    "eventTime": "",
+}
+
+EXCLUDED_DRONE_MEDIA_EVENT = {
+    "eventID": "TheWilds_2025_drone_20250703_parrotanafi_4_VID_001",
+    "parentEventID": "TheWilds_2025_drone_20250703_parrotanafi_4",
+    "eventType": "MediaEvent",
+    "eventDate": "2025-07-03",
+    "samplingProtocol": "Drone video VID_001; intentionally excluded from SAM3 and Tree-of-Life processing after review found no significant content.",
+    "sampleSizeValue": "0",
+    "sampleSizeUnit": "classified detection crops",
+    "locality": "The Wilds Conservation Center",
+    "locationID": "20250703_parrotanafi_4",
+    "decimalLatitude": "39.830678500000175",
+    "decimalLongitude": "-81.73834333333349",
+    "coordinateUncertaintyInMeters": "100",
+    "geodeticDatum": "WGS84",
+    "country": "United States",
+    "stateProvince": "Ohio",
+    "county": "Muskingum",
+    "datasetID": "TheWilds_2025_drones",
+    "datasetName": "The Wilds 2025 Multi-Modal Survey - Drone Imagery",
+    "recordedBy": "Drone field team",
+    "fieldNotes": "Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_4; video: VID_001; processingStatus: intentionally_excluded; reason: tiny video reviewed with no significant content; raw media: thewilds_20250630_20250703/drones/20250703_parrotanafi_4/video_files/VID_001.MP4",
+    "eventTime": "",
+}
+
+TAXA = {
+    "African Wild Dog": ("Lycaon pictus (Temminck, 1820)", "Animalia", "Chordata", "Mammalia", "Carnivora", "Canidae", "Lycaon", "pictus", "species"),
+    "Bactrian Camel": ("Camelus bactrianus Linnaeus, 1758", "Animalia", "Chordata", "Mammalia", "Artiodactyla", "Camelidae", "Camelus", "bactrianus", "species"),
+    "Bactrian Deer": ("Cervus hanglu bactrianus Lydekker, 1900", "Animalia", "Chordata", "Mammalia", "Artiodactyla", "Cervidae", "Cervus", "hanglu", "subspecies"),
+    "Banteng": ("Bos javanicus d'Alton, 1823", "Animalia", "Chordata", "Mammalia", "Artiodactyla", "Bovidae", "Bos", "javanicus", "species"),
+    "Bird": ("Aves", "Animalia", "Chordata", "Aves", "", "", "", "", "class"),
+    "Bison": ("Bison bison (Linnaeus, 1758)", "Animalia", "Chordata", "Mammalia", "Artiodactyla", "Bovidae", "Bison", "bison", "species"),
+    "Canada Goose": ("Branta canadensis (Linnaeus, 1758)", "Animalia", "Chordata", "Aves", "Anseriformes", "Anatidae", "Branta", "canadensis", "species"),
+    "Camel": ("Camelus", "Animalia", "Chordata", "Mammalia", "Artiodactyla", "Camelidae", "Camelus", "", "genus"),
+    "Coyote": ("Canis latrans Say, 1823", "Animalia", "Chordata", "Mammalia", "Carnivora", "Canidae", "Canis", "latrans", "species"),
+    "Donkey": ("Equus asinus Linnaeus, 1758", "Animalia", "Chordata", "Mammalia", "Perissodactyla", "Equidae", "Equus", "asinus", "species"),
+    "Eland": ("Taurotragus oryx (Pallas, 1766)", "Animalia", "Chordata", "Mammalia", "Artiodactyla", "Bovidae", "Taurotragus", "oryx", "species"),
+    "Elk": ("Cervus canadensis Erxleben, 1777", "Animalia", "Chordata", "Mammalia", "Artiodactyla", "Cervidae", "Cervus", "canadensis", "species"),
+    "Fish": ("Actinopterygii", "Animalia", "Chordata", "Actinopterygii", "", "", "", "", "class"),
+    "Fox": ("Vulpes", "Animalia", "Chordata", "Mammalia", "Carnivora", "Canidae", "Vulpes", "", "genus"),
+    "Fringe-eared Oryx": ("Oryx beisa callotis Thomas, 1892", "Animalia", "Chordata", "Mammalia", "Artiodactyla", "Bovidae", "Oryx", "beisa", "subspecies"),
+    "Giraffe": ("Giraffa camelopardalis (Linnaeus, 1758)", "Animalia", "Chordata", "Mammalia", "Artiodactyla", "Giraffidae", "Giraffa", "camelopardalis", "species"),
+    "Goat": ("Capra", "Animalia", "Chordata", "Mammalia", "Artiodactyla", "Bovidae", "Capra", "", "genus"),
+    "Greater One-horned Rhino": ("Rhinoceros unicornis Linnaeus, 1758", "Animalia", "Chordata", "Mammalia", "Perissodactyla", "Rhinocerotidae", "Rhinoceros", "unicornis", "species"),
+    "Grevy's Zebra": ("Equus grevyi Oustalet, 1882", "Animalia", "Chordata", "Mammalia", "Perissodactyla", "Equidae", "Equus", "grevyi", "species"),
+    "Hawk": ("Accipitridae", "Animalia", "Chordata", "Aves", "Accipitriformes", "Accipitridae", "", "", "family"),
+    "Heron": ("Ardeidae", "Animalia", "Chordata", "Aves", "Pelecaniformes", "Ardeidae", "", "", "family"),
+    "Human": ("Homo sapiens Linnaeus, 1758", "Animalia", "Chordata", "Mammalia", "Primates", "Hominidae", "Homo", "sapiens", "species"),
+    "Muskox": ("Ovibos moschatus (Zimmermann, 1780)", "Animalia", "Chordata", "Mammalia", "Artiodactyla", "Bovidae", "Ovibos", "moschatus", "species"),
+    "Osprey": ("Pandion haliaetus (Linnaeus, 1758)", "Animalia", "Chordata", "Aves", "Accipitriformes", "Pandionidae", "Pandion", "haliaetus", "species"),
+    "Ostrich": ("Struthio camelus Linnaeus, 1758", "Animalia", "Chordata", "Aves", "Struthioniformes", "Struthionidae", "Struthio", "camelus", "species"),
+    "Pere David's Deer": ("Elaphurus davidianus Milne-Edwards, 1866", "Animalia", "Chordata", "Mammalia", "Artiodactyla", "Cervidae", "Elaphurus", "davidianus", "species"),
+    "Persian Onager": ("Equus hemionus onager Boddaert, 1785", "Animalia", "Chordata", "Mammalia", "Perissodactyla", "Equidae", "Equus", "hemionus", "subspecies"),
+    "Przewalski's Horse": ("Equus ferus przewalskii Poliakov, 1881", "Animalia", "Chordata", "Mammalia", "Perissodactyla", "Equidae", "Equus", "ferus", "subspecies"),
+    "Scimitar-horned Oryx": ("Oryx dammah (Cretzschmar, 1827)", "Animalia", "Chordata", "Mammalia", "Artiodactyla", "Bovidae", "Oryx", "dammah", "species"),
+    "Sichuan Takin": ("Budorcas taxicolor tibetana Milne-Edwards, 1874", "Animalia", "Chordata", "Mammalia", "Artiodactyla", "Bovidae", "Budorcas", "taxicolor", "subspecies"),
+    "Southern White Rhino": ("Ceratotherium simum simum (Burchell, 1817)", "Animalia", "Chordata", "Mammalia", "Perissodactyla", "Rhinocerotidae", "Ceratotherium", "simum", "subspecies"),
+    "Waterfowl": ("Anatidae", "Animalia", "Chordata", "Aves", "Anseriformes", "Anatidae", "", "", "family"),
+    "White-tailed Deer": ("Odocoileus virginianus (Zimmermann, 1780)", "Animalia", "Chordata", "Mammalia", "Artiodactyla", "Cervidae", "Odocoileus", "virginianus", "species"),
+}
+
+ALIASES = {
+    "african wild dog": "African Wild Dog",
+    "antelope": "Eland",
+    "banteng": "Banteng",
+    "bird": "Bird",
+    "bison": "Bison",
+    "bos": "Banteng",
+    "budorcas": "Sichuan Takin",
+    "camel": "Camel",
+    "camelus": "Camel",
+    "canada goose": "Canada Goose",
+    "canis": "Coyote",
+    "cattle": "Banteng",
+    "ceratotherium": "Southern White Rhino",
+    "cervus": "Elk",
+    "coyote": "Coyote",
+    "deer": "White-tailed Deer",
+    "donkey": "Donkey",
+    "eland": "Eland",
+    "elaphurus": "Pere David's Deer",
+    "equus": "Przewalski's Horse",
+    "fox": "Fox",
+    "giraffa": "Giraffe",
+    "giraffe": "Giraffe",
+    "goat": "Goat",
+    "grevy's zebra": "Grevy's Zebra",
+    "hawk": "Hawk",
+    "heron": "Heron",
+    "horse": "Przewalski's Horse",
+    "human": "Human",
+    "lycaon": "African Wild Dog",
+    "elk": "Elk",
+    "milu deer": "Pere David's Deer",
+    "muskox": "Muskox",
+    "odocoileus": "White-tailed Deer",
+    "oryx": "Fringe-eared Oryx",
+    "osprey": "Osprey",
+    "ostrich": "Ostrich",
+    "pere david's deer": "Pere David's Deer",
+    "persian onager": "Persian Onager",
+    "person": "Human",
+    "przewalski's horse": "Przewalski's Horse",
+    "raptor": "Bird",
+    "rhinoceros": "Greater One-horned Rhino",
+    "sichuan takin": "Sichuan Takin",
+    "songbird": "Bird",
+    "struthio": "Ostrich",
+    "taurotragus": "Eland",
+    "waterfowl": "Waterfowl",
+    "white-tailed deer": "White-tailed Deer",
+    "wild dog": "African Wild Dog",
+    "zebra": "Grevy's Zebra",
+}
+
+
+def clean_label(label: str) -> str:
+    return (label or "").replace("Père", "Pere").strip()
+
+
+def taxon_for(label: str) -> dict[str, str]:
+    clean = clean_label(label)
+    canonical = ALIASES.get(clean.lower(), clean)
+    if canonical == "vehicle":
+        return {
+            "scientificName": "",
+            "kingdom": "",
+            "phylum": "",
+            "class": "",
+            "order": "",
+            "family": "",
+            "genus": "",
+            "specificEpithet": "",
+            "taxonRank": "",
+            "vernacularName": "vehicle",
+        }
+    sci, kingdom, phylum, klass, order, family, genus, epithet, rank = TAXA.get(
+        canonical, ("", "Animalia", "Chordata", "", "", "", "", "", "")
+    )
+    return {
+        "scientificName": sci,
+        "kingdom": kingdom,
+        "phylum": phylum,
+        "class": klass,
+        "order": order,
+        "family": family,
+        "genus": genus,
+        "specificEpithet": epithet,
+        "taxonRank": rank,
+        "vernacularName": canonical,
+    }
+
+
+def ensure_clean_output(force: bool) -> None:
+    OUT_DIR.mkdir(parents=True, exist_ok=True)
+    (OUT_DIR / "scripts").mkdir(parents=True, exist_ok=True)
+    outputs = [
+        OUT_DIR / "camera_traps",
+        OUT_DIR / "bioacoustics_complete",
+        OUT_DIR / "drones",
+        OUT_DIR / "citizen_science",
+        OUT_DIR / "text_data",
+        OUT_DIR / "events_all_modalities.csv",
+        OUT_DIR / "occurrences_all_modalities.csv",
+        OUT_DIR / "QC_SUMMARY.md",
+    ]
+    existing = [p for p in outputs if p.exists()]
+    if existing and not force:
+        names = "\n".join(str(p) for p in existing)
+        raise SystemExit(f"Refusing to overwrite existing outputs. Re-run with --force.\n{names}")
+    if force:
+        for p in existing:
+            if p.is_dir():
+                shutil.rmtree(p)
+            else:
+                p.unlink()
+
+
+def write_csv(path: Path, rows: list[dict], fields: list[str]) -> None:
+    path.parent.mkdir(parents=True, exist_ok=True)
+    with path.open("w", newline="", encoding="utf-8") as f:
+        writer = csv.DictWriter(f, fieldnames=fields, extrasaction="ignore")
+        writer.writeheader()
+        writer.writerows(rows)
+
+
+def read_csv(path: Path) -> list[dict]:
+    with path.open(newline="", encoding="utf-8", errors="replace") as f:
+        return list(csv.DictReader(f))
+
+
+DATE_SENTINELS = {"", "nan", "nat", "none", "null", "n/a", "na", "-1", "1970-01-01"}
+
+# The only original-video locations used as a fallback for invalid drone CSV coordinates.
+# Values are read directly from each MP4's QuickTime ISO-6709 metadata.
+DRONE_ORIGINAL_VIDEO_PATHS = {
+    ("20250811_parrot_01", "P5180524"): BASE_DIR / "thewilds_20250804_20250828/drones/20250811_parrot_01/videos/P5180524.MP4",
+    ("20250811_parrot_01", "P5190525"): BASE_DIR / "thewilds_20250804_20250828/drones/20250811_parrot_01/videos/P5190525.MP4",
+    ("20250811_parrot_01", "P5190526"): BASE_DIR / "thewilds_20250804_20250828/drones/20250811_parrot_01/videos/P5190526.MP4",
+    ("20250703_parrotanafi_5", "VID_001"): BASE_DIR / "thewilds_20250630_20250703/drones/20250703_parrotanafi_5/videos/VID_001.MP4",
+    ("20250703_parrotanafi_5", "VID_002"): BASE_DIR / "thewilds_20250630_20250703/drones/20250703_parrotanafi_5/videos/VID_002.MP4",
+}
+QUICKTIME_ISO6709_RE = re.compile(
+    rb"([+-]\d{1,3}(?:\.\d+)?)([+-]\d{1,3}(?:\.\d+)?)(?:[+-]\d+(?:\.\d+)?)?/"
+)
+DRONE_VIDEO_LOCATION_CACHE: dict[tuple[str, str], tuple[str, str] | None] = {}
+
+
+def valid_wgs84_coordinates(latitude: str, longitude: str) -> bool:
+    """Return whether a pair is finite and inside WGS84 latitude/longitude bounds."""
+    try:
+        lat = float(latitude)
+        lon = float(longitude)
+    except (TypeError, ValueError):
+        return False
+    return math.isfinite(lat) and math.isfinite(lon) and -90 <= lat <= 90 and -180 <= lon <= 180
+
+
+def valid_wilds_video_location(latitude: str, longitude: str) -> bool:
+    """Limit binary QuickTime matches to the documented study area before use."""
+    if not valid_wgs84_coordinates(latitude, longitude):
+        return False
+    lat = float(latitude)
+    lon = float(longitude)
+    return 39.6 <= lat <= 40.1 and -82.1 <= lon <= -81.4
+
+
+def original_video_location(flight: str, video_stem: str) -> tuple[str, str] | None:
+    """Read one valid QuickTime ISO-6709 location from a whitelisted original MP4."""
+    key = (flight, video_stem)
+    if key in DRONE_VIDEO_LOCATION_CACHE:
+        return DRONE_VIDEO_LOCATION_CACHE[key]
+
+    path = DRONE_ORIGINAL_VIDEO_PATHS.get(key)
+    if path is None or not path.is_file():
+        DRONE_VIDEO_LOCATION_CACHE[key] = None
+        return None
+
+    size = path.stat().st_size
+    window = 16 * 1024 * 1024
+    with path.open("rb") as handle:
+        chunks = [handle.read(window)]
+        if size > window:
+            handle.seek(max(0, size - window))
+            chunks.append(handle.read(window))
+
+    locations = {
+        (match.group(1).decode("ascii").lstrip("+"), match.group(2).decode("ascii").lstrip("+"))
+        for chunk in chunks
+        for match in QUICKTIME_ISO6709_RE.finditer(chunk)
+        if valid_wilds_video_location(
+            match.group(1).decode("ascii"), match.group(2).decode("ascii")
+        )
+    }
+    # A video-level fallback is defensible only when its metadata contains one location.
+    location = next(iter(locations)) if len(locations) == 1 else None
+    DRONE_VIDEO_LOCATION_CACHE[key] = location
+    return location
+
+
+def resolve_drone_occurrence_coordinates(row: dict, flight: str, video_stem: str) -> tuple[str, str, str, str]:
+    """Resolve occurrence coordinates without allowing invalid CSV values into DwC."""
+    if row.get("drone_georef_status") == "ok":
+        lat = row.get("animal_lat_projected", "")
+        lon = row.get("animal_lon_projected", "")
+        if valid_wgs84_coordinates(lat, lon):
+            return lat, lon, "50", "projected_animal_position"
+
+    lat = row.get("latitude_y", "")
+    lon = row.get("longitude_y", "")
+    if valid_wgs84_coordinates(lat, lon):
+        return lat, lon, "100", "frame_or_drone_telemetry"
+
+    fallback = original_video_location(flight, video_stem)
+    if fallback:
+        return fallback[0], fallback[1], "500", "original_video_quicktime_location_static"
+    return "", "", "", "unavailable_no_valid_source_coordinate"
+
+
+def resolve_drone_event_coordinates(row: dict, flight: str, video_stem: str) -> tuple[str, str, str, str]:
+    """Resolve a video Event position from telemetry or a static original-video fallback."""
+    lat = row.get("latitude_y", "")
+    lon = row.get("longitude_y", "")
+    if valid_wgs84_coordinates(lat, lon):
+        return lat, lon, "100", "frame_or_drone_telemetry"
+
+    fallback = original_video_location(flight, video_stem)
+    if fallback:
+        return fallback[0], fallback[1], "500", "original_video_quicktime_location_static"
+    return "", "", "", "unavailable_no_valid_source_coordinate"
+
+
+def drone_coordinate_generalization(source: str) -> str:
+    if source == "projected_animal_position":
+        return "Approximate animal ground-position projection from drone telemetry and bbox bottom-center."
+    if source == "frame_or_drone_telemetry":
+        return "Frame/drone telemetry coordinate; not a projected animal position."
+    if source == "original_video_quicktime_location_static":
+        return "Static video-level location read from original-video QuickTime ISO-6709 metadata after the CSV coordinate failed WGS84 validation; it is not a frame- or animal-level position."
+    return "No valid WGS84 coordinate was available from the geolocation CSV or original-video metadata."
+
+
+DRONE_VIDEO_MEDIA_CACHE: dict[tuple[str, str, str, str], str] = {}
+
+
+def resolve_drone_associated_media(row: dict, collection: str, flight: str, video_stem: str) -> tuple[str, str]:
+    """Keep an extant extracted frame, otherwise link to its retained source video."""
+    frame_path = row.get("frame_path", "")
+    if frame_path and Path(frame_path).is_file():
+        return frame_path, "extracted_frame"
+
+    key = (collection, flight, row.get("subdir", ""), video_stem)
+    if key not in DRONE_VIDEO_MEDIA_CACHE:
+        video_dir = BASE_DIR / collection / "drones" / flight / row.get("subdir", "")
+        candidates = sorted(item for item in video_dir.glob(video_stem + ".*") if item.is_file())
+        DRONE_VIDEO_MEDIA_CACHE[key] = str(candidates[0]) if len(candidates) == 1 else ""
+    video_path = DRONE_VIDEO_MEDIA_CACHE[key]
+    if video_path:
+        return video_path, "original_video_fallback_missing_scratch_frame"
+    return "", "unavailable_no_retained_media"
+
+
+def split_datetime(value: str) -> tuple[str, str]:
+    """Return an ISO date/time only for a parseable timestamp."""
+    value = (value or "").strip()
+    if value.lower() in DATE_SENTINELS:
+        return "", ""
+    try:
+        parsed = datetime.fromisoformat(value.replace("Z", "+00:00"))
+    except ValueError:
+        return "", ""
+    return parsed.date().isoformat(), parsed.time().replace(microsecond=0).isoformat()
+
+
+def date_from_identifier(identifier: str) -> str:
+    """Extract and validate a YYYYMMDD date token from a flight identifier."""
+    match = re.search(r"(?<!\d)(20\d{6})(?!\d)", identifier or "")
+    if not match:
+        return ""
+    try:
+        return date.fromisoformat(
+            f"{match.group(1)[:4]}-{match.group(1)[4:6]}-{match.group(1)[6:]}"
+        ).isoformat()
+    except ValueError:
+        return ""
+
+
+def load_camera_timestamp_map() -> dict[str, dict]:
+    path = BASE_DIR / "processing_output/detections/drone_camera_analysis/camera_time_mapping/camera_image_timestamp_map.csv"
+    corrections = {}
+    for row in read_csv(path):
+        image_path = row.get("image_path", "")
+        if image_path and row.get("corrected_timestamp"):
+            corrections[image_path] = row
+    return corrections
+
+
+def load_camera_geolocation_config() -> list[dict]:
+    path = BASE_DIR / "processing_output/detections/camera_traps/methods_paper/annotations/sam3/treeoflife_sam3/geolocation/camera_geolocation_config.csv"
+    return [row for row in read_csv(path) if row.get("enabled", "1") not in {"0", "false", "False"}]
+
+
+def choose_camera_config(configs: list[dict], collection: str, camera_id: str, frame_path: str) -> dict:
+    candidates = [
+        row for row in configs
+        if row.get("collection") == collection and row.get("camera_id") == camera_id
+    ]
+    matching = [
+        row for row in candidates
+        if not row.get("path_contains") or row.get("path_contains") in frame_path
+    ]
+    if not matching:
+        return {}
+    return max(matching, key=lambda row: len(row.get("path_contains", "")))
+
+
+def camera_csv_parts(csv_path: Path) -> tuple[str, str, str, str]:
+    collection = csv_path.parent.parent.name
+    camera_id = csv_path.parent.name
+    if collection == "thewilds_20250630_20250703":
+        event_range = "2025-06-30/2025-07-03"
+    elif collection == "thewilds_20250804_20250828":
+        event_range = "2025-08-04/2025-08-28"
+    else:
+        event_range = ""
+    return collection, camera_id, collection.replace("thewilds_", ""), event_range
+
+
+def site_from_camera(camera_id: str) -> str:
+    return camera_id.replace("_", "-").split("-", 1)[0]
+
+
+def taxon_from_treeoflife(row: dict) -> dict[str, str]:
+    scientific = (
+        row.get("top1_scientific_name")
+        or row.get("top1_species")
+        or row.get("top1_genus")
+        or row.get("top1_family")
+        or row.get("top1_order")
+        or row.get("top1_class")
+        or row.get("top1_common_name")
+        or ""
+    )
+    genus = row.get("top1_genus", "")
+    epithet = ""
+    if row.get("top1_species"):
+        parts = row["top1_species"].split()
+        if len(parts) >= 2:
+            epithet = parts[-1]
+    elif scientific and len(scientific.split()) >= 2:
+        epithet = scientific.split()[1]
+    return {
+        "scientificName": scientific,
+        "kingdom": row.get("top1_kingdom", ""),
+        "phylum": row.get("top1_phylum", ""),
+        "class": row.get("top1_class", ""),
+        "order": row.get("top1_order", ""),
+        "family": row.get("top1_family", ""),
+        "genus": genus,
+        "specificEpithet": epithet,
+        "taxonRank": row.get("top1_rank") or row.get("rank", ""),
+        "vernacularName": row.get("top1_common_name", ""),
+    }
+
+
+def generate_camera_traps() -> tuple[int, int]:
+    det_dir = BASE_DIR / "processing_output/detections/camera_traps/methods_paper/annotations/sam3/treeoflife_sam3/geolocation/geolocated_predictions"
+    files = sorted(det_dir.glob("*/*/*.geolocated.csv"))
+    corrections = load_camera_timestamp_map()
+    geo_configs = load_camera_geolocation_config()
+    events = [{
+        "eventID": "TheWilds_2025_camera_traps",
+        "parentEventID": "TheWilds_2025",
+        "eventType": "Project",
+        "eventDate": "2025-06-30/2025-08-28",
+        "samplingProtocol": "Camera trap monitoring; detections by SAM3, Tree-of-Life classification by BioCLIP, animal position estimated from camera GPS/FOV/bearing and detection boxes.",
+        "sampleSizeValue": str(len(files)),
+        "sampleSizeUnit": "camera Tree-of-Life geolocation CSVs",
+        "locality": "The Wilds Conservation Center, Muskingum County, Ohio, USA",
+        "locationID": "",
+        "decimalLatitude": "39.83",
+        "decimalLongitude": "-81.74",
+        "coordinateUncertaintyInMeters": "5000",
+        "geodeticDatum": "WGS84",
+        "country": "United States",
+        "stateProvince": "Ohio",
+        "county": "Muskingum",
+        "datasetID": "TheWilds_2025_camera_traps",
+        "datasetName": "The Wilds 2025 Multi-Modal Survey - Camera Traps",
+        "recordedBy": "Field team",
+        "fieldNotes": "Regenerated from SAM3 + BioCLIP Tree-of-Life detections using the latest timestamp mapping and geolocation outputs.",
+        "eventTime": "",
+    }]
+    occurrences = []
+    occ_num = 0
+    for csv_path in files:
+        collection, site_cam, campaign, event_range = camera_csv_parts(csv_path)
+        site = site_from_camera(site_cam)
+        rows = read_csv(csv_path)
+        first_cfg = choose_camera_config(geo_configs, collection, site_cam, rows[0].get("frame_path", "") if rows else "")
+        lat = first_cfg.get("camera_lat") or SITE_GPS.get(site, ("39.83", "-81.74"))[0]
+        lon = first_cfg.get("camera_lon") or SITE_GPS.get(site, ("39.83", "-81.74"))[1]
+        site_name = SITE_NAMES.get(site, site)
+        event_id = f"TheWilds_2025_CT_deployment_{site_cam}_{campaign}"
+        events.append({
+            "eventID": event_id,
+            "parentEventID": "TheWilds_2025_camera_traps",
+            "eventType": "SensorDeployment",
+            "eventDate": event_range,
+            "samplingProtocol": f"Trail camera ({site_cam}); SAM3 detections with BioCLIP Tree-of-Life labels and geolocation estimates",
+            "sampleSizeValue": str(len(rows)),
+            "sampleSizeUnit": "detections",
+            "locality": f"The Wilds - {site_name}",
+            "locationID": site,
+            "decimalLatitude": lat,
+            "decimalLongitude": lon,
+            "coordinateUncertaintyInMeters": "5",
+            "geodeticDatum": "WGS84",
+            "country": "United States",
+            "stateProvince": "Ohio",
+            "county": "Muskingum",
+            "datasetID": "TheWilds_2025_camera_traps",
+            "datasetName": "The Wilds 2025 Multi-Modal Survey - Camera Traps",
+            "recordedBy": "Field team",
+            "fieldNotes": f"Camera: {site_cam}; Campaign: {campaign}; detections: {len(rows)}; geolocation config source: {first_cfg.get('config_source', '')}",
+            "eventTime": "",
+        })
+        for row in rows:
+            occ_num += 1
+            frame_path = row.get("frame_path", "")
+            time_row = corrections.get(frame_path, {})
+            taxon = taxon_from_treeoflife(row)
+            timestamp = time_row.get("corrected_timestamp") or row.get("date_time", "")
+            event_date, event_time = split_datetime(timestamp)
+            occ_lat = row.get("animal_lat") or row.get("camera_lat") or lat
+            occ_lon = row.get("animal_lon") or row.get("camera_lon") or lon
+            bbox = f"[{row.get('bbox_x','')},{row.get('bbox_y','')},{row.get('bbox_width','')},{row.get('bbox_height','')}]"
+            associated_media, media_source = resolve_drone_associated_media(
+                row, collection, flight, video_stem
+            )
+            rank = row.get("rank", "")
+            score = row.get("top1_score", "")
+            source_label = row.get("source_label", "")
+            correction_note = "; ".join(part for part in [
+                f"timestamp_method: {time_row.get('correction_method', '')}" if time_row else "",
+                f"timestamp_confidence: {time_row.get('correction_confidence', '')}" if time_row else "",
+                f"geolocation_status: {row.get('geolocation_status', '')}",
+                f"geolocation_source: {row.get('geolocation_config_source', '')}",
+            ] if part)
+            occurrences.append({
+                "occurrenceID": f"TheWilds_2025_CT_{occ_num:06d}",
+                "eventID": event_id,
+                "catalogNumber": f"CT_{site_cam}_image{row.get('image_id','')}_ann{row.get('annotation_id','')}",
+                **taxon,
+                "occurrenceStatus": "present",
+                "individualCount": "1",
+                "basisOfRecord": "MachineObservation",
+                "recordedBy": f"Trail camera ({site_cam})",
+                "identifiedBy": "SAM3 + BioCLIP Tree of Life",
+                "identificationVerificationStatus": "unverified",
+                "identificationRemarks": f"Tree-of-Life rank: {rank}; top1 score: {score}; detector label: {source_label}; bbox xywh(px): {bbox}",
+                "eventDate": event_date,
+                "eventTime": event_time,
+                "decimalLatitude": occ_lat,
+                "decimalLongitude": occ_lon,
+                "coordinateUncertaintyInMeters": "50",
+                "geodeticDatum": "WGS84",
+                "locality": f"The Wilds - {time_row.get('location') or site_name}",
+                "locationID": site,
+                "country": "United States",
+                "stateProvince": "Ohio",
+                "county": "Muskingum",
+                "establishmentMeans": "managed",
+                "degreeOfEstablishment": "captive",
+                "occurrenceRemarks": f"Estimated distance: {row.get('animal_distance_m','')}m; estimated bearing: {row.get('animal_absolute_bearing_deg','')} deg; {correction_note}",
+                "associatedMedia": frame_path,
+                "datasetID": "TheWilds_2025_camera_traps",
+                "datasetName": "The Wilds 2025 Multi-Modal Survey - Camera Traps",
+                "informationWithheld": "",
+                "dataGeneralizations": "Animal coordinates are model-derived estimates from camera position, camera bearing, HFOV, bbox height, and assumed target height; not GPS collars or direct observations.",
+            })
+    out = OUT_DIR / "camera_traps"
+    write_csv(out / "darwin_core_events_camera_traps.csv", events, EVENT_FIELDS)
+    write_csv(out / "darwin_core_occurrences_camera_traps.csv", occurrences, OCC_FIELDS)
+    return len(events), len(occurrences)
+
+
+def build_birdnet_scientific_map() -> dict[str, str]:
+    mapping = {}
+    roots = [
+        BASE_DIR / "processing_output/detections/acoustics/BirdnetCSV_SunriseSunset",
+        BASE_DIR / "processing_output/detections/acoustics/BirdNet_EveryHour",
+    ]
+    for root in roots:
+        for path in root.glob("*.BirdNET.results.csv"):
+            for row in read_csv(path):
+                common = (row.get("Common name") or row.get("Common Name") or "").strip()
+                sci = (row.get("Scientific name") or row.get("Scientific Name") or "").strip()
+                if common and sci:
+                    mapping.setdefault(common, sci)
+    return mapping
+
+
+BIOACOUSTIC_FILE_ID_MAP = {
+    "SM001": "SM01",
+    "SM002": "SM02",
+    "2MM18157": "SM05",
+    "2MM18027": "SM06",
+}
+BIOACOUSTIC_RESULT_SOURCES = [
+    (BASE_DIR / "processing_output/detections/acoustics/BirdNet_EveryHour", "every_hour"),
+    (BASE_DIR / "BioacousticAnalysis/BirdnetCSV_SunriseSunset", "sunrise_sunset"),
+]
+BIOACOUSTIC_RESULT_NAME_RE = re.compile(r"^([A-Za-z0-9]+)_(\d{8})_(\d{6})\.BirdNET\.results\.csv$")
+
+
+def add_header_only_birdnet_events(events: list[dict[str, str]]) -> int:
+    """Represent processed recordings with empty BirdNET result files as Events."""
+    existing_ids = {row["eventID"] for row in events}
+    deployments = []
+    recording_protocols = {}
+    for row in events:
+        if row.get("eventType") == "SensorDeployment":
+            match = re.search(r"Recorder: (SM\d+)", row.get("fieldNotes", ""))
+            if match:
+                start, end = row["eventDate"].split("/")
+                deployments.append((match.group(1), date.fromisoformat(start), date.fromisoformat(end), row))
+        elif row.get("eventType") == "AudioRecording":
+            recording_protocols.setdefault(row.get("parentEventID", ""), row.get("samplingProtocol", ""))
+
+    additions = []
+    for result_dir, schedule_type in BIOACOUSTIC_RESULT_SOURCES:
+        for path in sorted(result_dir.glob("*.BirdNET.results.csv")):
+            if read_csv(path):
+                continue
+            match = BIOACOUSTIC_RESULT_NAME_RE.match(path.name)
+            if not match:
+                continue
+            raw_recorder, ymd, hms = match.groups()
+            recorder = BIOACOUSTIC_FILE_ID_MAP.get(raw_recorder, raw_recorder)
+            recording_date = date.fromisoformat(f"{ymd[:4]}-{ymd[4:6]}-{ymd[6:]}")
+            deployment = next(
+                (row for rec, start, end, row in deployments
+                 if rec == recorder and start <= recording_date <= end),
+                None,
+            )
+            if deployment is None:
+                raise ValueError(f"No deployment for header-only BirdNET result: {path}")
+            file_id = f"{recorder}_{ymd}_{hms}"
+            event_id = f"TheWilds_2025_ARU_{file_id}"
+            if event_id in existing_ids:
+                continue
+            protocol = recording_protocols.get(deployment["eventID"], deployment["samplingProtocol"])
+            additions.append({
+                "eventID": event_id,
+                "parentEventID": deployment["eventID"],
+                "eventType": "AudioRecording",
+                "eventDate": recording_date.isoformat(),
+                "samplingProtocol": protocol,
+                "sampleSizeValue": "1",
+                "sampleSizeUnit": "WAV recording",
+                "locality": deployment["locality"],
+                "locationID": deployment["locationID"],
+                "decimalLatitude": deployment["decimalLatitude"],
+                "decimalLongitude": deployment["decimalLongitude"],
+                "coordinateUncertaintyInMeters": deployment["coordinateUncertaintyInMeters"],
+                "geodeticDatum": deployment["geodeticDatum"],
+                "country": deployment["country"],
+                "stateProvince": deployment["stateProvince"],
+                "county": deployment["county"],
+                "datasetID": deployment["datasetID"],
+                "datasetName": deployment["datasetName"],
+                "recordedBy": deployment["recordedBy"],
+                "fieldNotes": (
+                    f"Recording: {file_id}.wav; Recorder: {recorder}; "
+                    "Total BirdNET detections: 0; "
+                    f"Schedule type: {schedule_type}; "
+                    f"Provenance: header-only BirdNET result {path.name}."
+                ),
+                "eventTime": f"{hms[:2]}:{hms[2:4]}:{hms[4:]}",
+            })
+            existing_ids.add(event_id)
+    if len(additions) != 64:
+        raise ValueError(f"Expected 64 header-only BirdNET Events, found {len(additions)}")
+    events.extend(sorted(additions, key=lambda row: row["eventID"]))
+    return len(additions)
+
+
+def patch_bioacoustics() -> tuple[int, int]:
+    src = BASE_DIR / "processing_output/darwin_core/bioacoustics_complete"
+    out = OUT_DIR / "bioacoustics_complete"
+    out.mkdir(parents=True, exist_ok=True)
+    events = read_csv(src / "darwin_core_events_bioacoustics_complete.csv")
+    occ = read_csv(src / "darwin_core_occurrences_bioacoustics_complete.csv")
+    added_empty_events = add_header_only_birdnet_events(events)
+    mapping = build_birdnet_scientific_map()
+    for row in occ:
+        sci = row.get("scientificName") or mapping.get(row.get("vernacularName", ""), "")
+        row["scientificName"] = sci
+        if sci:
+            parts = sci.split()
+            row["kingdom"] = row.get("kingdom") or "Animalia"
+            row["phylum"] = row.get("phylum") or "Chordata"
+            row["class"] = row.get("class") or "Aves"
+            row["genus"] = row.get("genus") or parts[0]
+            row["specificEpithet"] = row.get("specificEpithet") or (parts[1] if len(parts) > 1 else "")
+            row["taxonRank"] = row.get("taxonRank") or ("species" if len(parts) > 1 else "")
+    write_csv(out / "darwin_core_events_bioacoustics_complete.csv", events, list(events[0].keys()))
+    write_csv(out / "darwin_core_occurrences_bioacoustics_complete.csv", occ, list(occ[0].keys()))
+    results = src / "RESULTS.md"
+    if results.exists():
+        summary = results.read_text(encoding="utf-8")
+        summary = summary.replace("**2,307 recordings**", "**2,371 recordings**")
+        summary = summary.replace("| Campaign 1 | Jun 30 – Jul 3 | 300 |", "| Campaign 1 | Jun 30 – Jul 3 | 311 |")
+        summary = summary.replace("| Campaign 2 | Aug 4 – Aug 28 | 2,007 |", "| Campaign 2 | Aug 4 – Aug 28 | 2,060 |")
+        summary = summary.replace("| TW02 | Nomad Ridge Shelter | Aug 4–28 | 561 |", "| TW02 | Nomad Ridge Shelter | Aug 4–28 | 581 |")
+        summary = summary.replace("| TW06 | Zebra Shelter | Jun 30–Aug 28 | 550 |", "| TW06 | Zebra Shelter | Jun 30–Aug 28 | 585 |")
+        summary = summary.replace("| TW08 | Fence with Vegetation | Jun 30–Aug 28 | 410 |", "| TW08 | Fence with Vegetation | Jun 30–Aug 28 | 419 |")
+        summary += "\n\n## Recording Completeness Repair\n\nThe corrected snapshot includes 64 processed recordings whose BirdNET result CSVs contain only a header and therefore have zero detections. They are represented as `AudioRecording` Events with no Occurrences.\n"
+        (out / "RESULTS.md").write_text(summary, encoding="utf-8")
+    return len(events), len(occ)
+
+
+def generate_drones() -> tuple[int, int]:
+    det_dir = BASE_DIR / "processing_output/detections/drones/geolocation/geolocated_predictions"
+    files = sorted(det_dir.glob("*.drone_geolocated.csv"))
+    if not files:
+        raise FileNotFoundError(f"No drone geolocated Tree-of-Life prediction CSVs found under {det_dir}")
+
+    project_event = {
+        "eventID": "TheWilds_2025_drones",
+        "parentEventID": "TheWilds_2025",
+        "eventType": "Project",
+        "eventDate": "2025-07-01/2025-09-08",
+        "samplingProtocol": "Drone video surveys; detections by SAM3, Tree-of-Life classification by BioCLIP, and first-pass animal ground-position projection from drone telemetry.",
+        "sampleSizeValue": str(len(files)),
+        "sampleSizeUnit": "drone video geolocated Tree-of-Life prediction CSVs",
+        "locality": "The Wilds Conservation Center, Muskingum County, Ohio, USA",
+        "locationID": "",
+        "decimalLatitude": "39.83",
+        "decimalLongitude": "-81.74",
+        "coordinateUncertaintyInMeters": "5000",
+        "geodeticDatum": "WGS84",
+        "country": "United States",
+        "stateProvince": "Ohio",
+        "county": "Muskingum",
+        "datasetID": "TheWilds_2025_drones",
+        "datasetName": "The Wilds 2025 Multi-Modal Survey - Drone Imagery",
+        "recordedBy": "Drone field team",
+        "fieldNotes": "38 videos regenerated from per-video SAM3 + BioCLIP Tree-of-Life prediction CSVs with approximate animal ground-position projections where possible; VID_001 from 20250703_parrotanafi_4 is retained as an intentionally excluded raw-media Event.",
+        "eventTime": "",
+    }
+    events = [project_event]
+    flight_events: dict[str, dict] = {}
+    occurrences = []
+    occ_num = 0
+    for path in files:
+        rows = read_csv(path)
+        if not rows:
+            continue
+
+        first = rows[0]
+        collection = first.get("collection", "")
+        flight = first.get("flight", "")
+        video_stem = first.get("video_stem", "") or path.stem.split("_")[-1]
+        fallback_date = date_from_identifier(flight)
+        date_values = sorted(
+            {
+                parsed_date
+                for row in rows
+                if (parsed_date := split_datetime(row.get("date_time", ""))[0])
+            }
+        )
+        event_date = (
+            date_values[0]
+            if len(date_values) == 1
+            else f"{date_values[0]}/{date_values[-1]}" if date_values else fallback_date
+        )
+        event_date_derived = not date_values and bool(fallback_date)
+        event_times = sorted(
+            {
+                parsed_time
+                for row in rows
+                if (parsed_time := split_datetime(row.get("date_time", ""))[1])
+            }
+        )
+        event_time = (
+            event_times[0]
+            if len(event_times) == 1
+            else f"{event_times[0]}/{event_times[-1]}" if event_times else ""
+        )
+        event_lat, event_lon, event_uncertainty, event_coordinate_source = resolve_drone_event_coordinates(
+            first, flight, video_stem
+        )
+        # A static location is valid for a single video, but not necessarily for a flight
+        # with multiple videos; keep the parent Event blank in that situation.
+        if event_coordinate_source == "frame_or_drone_telemetry":
+            flight_lat, flight_lon, flight_uncertainty = event_lat, event_lon, event_uncertainty
+        else:
+            flight_lat = flight_lon = flight_uncertainty = ""
+        flight_event_id = f"TheWilds_2025_drone_{flight}"
+        video_event_id = f"TheWilds_2025_drone_{flight}_{video_stem}"
+
+        if flight and flight_event_id not in flight_events:
+            flight_events[flight_event_id] = {
+                "eventID": flight_event_id,
+                "parentEventID": "TheWilds_2025_drones",
+                "eventType": "SensorDeployment",
+                "eventDate": event_date,
+                "samplingProtocol": f"Drone flight/session {flight}; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available",
+                "sampleSizeValue": "",
+                "sampleSizeUnit": "detections",
+                "locality": "The Wilds Conservation Center",
+                "locationID": flight,
+                "decimalLatitude": flight_lat,
+                "decimalLongitude": flight_lon,
+                "coordinateUncertaintyInMeters": flight_uncertainty,
+                "geodeticDatum": "WGS84",
+                "country": "United States",
+                "stateProvince": "Ohio",
+                "county": "Muskingum",
+                "datasetID": "TheWilds_2025_drones",
+                "datasetName": "The Wilds 2025 Multi-Modal Survey - Drone Imagery",
+                "recordedBy": "Drone field team",
+                "fieldNotes": f"Collection: {collection}; flight/session: {flight}; coordinate_source={event_coordinate_source}",
+                "eventTime": "",
+            }
+
+        events.append({
+            "eventID": video_event_id,
+            "parentEventID": flight_event_id,
+            "eventType": "MediaEvent",
+            "eventDate": event_date,
+            "samplingProtocol": f"Drone video {video_stem}; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available",
+            "sampleSizeValue": str(len(rows)),
+            "sampleSizeUnit": "classified detection crops",
+            "locality": "The Wilds Conservation Center",
+            "locationID": flight,
+            "decimalLatitude": event_lat,
+            "decimalLongitude": event_lon,
+            "coordinateUncertaintyInMeters": event_uncertainty,
+            "geodeticDatum": "WGS84",
+            "country": "United States",
+            "stateProvince": "Ohio",
+            "county": "Muskingum",
+            "datasetID": "TheWilds_2025_drones",
+            "datasetName": "The Wilds 2025 Multi-Modal Survey - Drone Imagery",
+            "recordedBy": "Drone field team",
+            "fieldNotes": f"Collection: {collection}; flight/session: {flight}; video: {video_stem}; detections: {len(rows)}; source CSV: {path.name}; coordinate_source={event_coordinate_source}",
+            "eventTime": event_time,
+        })
+
+        for row in rows:
+            occ_num += 1
+            taxon = taxon_from_treeoflife(row)
+            event_date, event_time = split_datetime(row.get("date_time", ""))
+            event_date_derived_for_row = not event_date and bool(fallback_date)
+            if event_date_derived_for_row:
+                event_date = fallback_date
+            lat, lon, coord_uncertainty, coordinate_source = resolve_drone_occurrence_coordinates(
+                row, flight, video_stem
+            )
+            coordinate_withheld = (
+                "Coordinate unavailable: no valid WGS84 geolocation CSV value or original-video metadata location."
+                if coordinate_source == "unavailable_no_valid_source_coordinate" else ""
+            )
+            information_withheld = "; ".join(part for part in [
+                "Intra-day timestamp unavailable; eventDate derived from flight identifier."
+                if event_date_derived_for_row else "",
+                coordinate_withheld,
+            ] if part)
+            bbox = f"[{row.get('bbox_x','')},{row.get('bbox_y','')},{row.get('bbox_width','')},{row.get('bbox_height','')}]"
+            rank = row.get("rank", "")
+            score = row.get("top1_score", "")
+            source_label = row.get("source_label", "")
+            occurrences.append({
+                "occurrenceID": f"TheWilds_2025_drone_{occ_num:06d}",
+                "eventID": video_event_id,
+                "catalogNumber": f"drone_{flight}_{video_stem}_image{row.get('image_id','')}_ann{row.get('annotation_id','')}",
+                **taxon,
+                "occurrenceStatus": "present",
+                "individualCount": "1",
+                "basisOfRecord": "MachineObservation",
+                "recordedBy": "Drone imagery",
+                "identifiedBy": "SAM3 + BioCLIP Tree of Life",
+                "identificationVerificationStatus": "unverified",
+                "identificationRemarks": f"Tree-of-Life rank: {rank}; top1 score: {score}; detector label: {source_label}; bbox xywh(px): {bbox}; altitude_m: {row.get('altitude_m','')}",
+                "eventDate": event_date,
+                "eventTime": event_time,
+                "decimalLatitude": lat,
+                "decimalLongitude": lon,
+                "coordinateUncertaintyInMeters": coord_uncertainty,
+                "geodeticDatum": "WGS84",
+                "locality": "The Wilds Conservation Center",
+                "locationID": flight,
+                "country": "United States",
+                "stateProvince": "Ohio",
+                "county": "Muskingum",
+                "establishmentMeans": "managed",
+                "degreeOfEstablishment": "captive",
+                "occurrenceRemarks": "; ".join(part for part in [
+                    "Frame-level drone detection crop",
+                    "event_date_source=flight_identifier" if event_date_derived_for_row else "event_date_source=source_datetime",
+                    f"source={row.get('source','')}",
+                    f"mission_id={row.get('mission_id','')}",
+                    f"detection_id={row.get('detection_id','')}",
+                    f"drone_georef_status={row.get('drone_georef_status','')}",
+                    f"coordinate_source={coordinate_source}",
+                    f"associated_media_source={media_source}",
+                    f"projected_ground_range_m={row.get('animal_ground_range_m','')}" if row.get("animal_ground_range_m") else "",
+                    f"projected_bearing_deg={row.get('animal_bearing_deg','')}" if row.get("animal_bearing_deg") else "",
+                    f"drone_altitude_used_m={row.get('drone_altitude_used_m','')}" if row.get("drone_altitude_used_m") else "",
+                    f"drone_pitch_source={row.get('drone_pitch_source','')}" if row.get("drone_pitch_source") else "",
+                ] if part),
+                "associatedMedia": associated_media,
+                "datasetID": "TheWilds_2025_drones",
+                "datasetName": "The Wilds 2025 Multi-Modal Survey - Drone Imagery",
+                "informationWithheld": information_withheld,
+                "dataGeneralizations": drone_coordinate_generalization(coordinate_source),
+            })
+    # Represent the reviewed-but-intentionally-unprocessed raw video explicitly.
+    if not any(row.get("eventID") == EXCLUDED_DRONE_MEDIA_EVENT["eventID"] for row in events):
+        events.append(dict(EXCLUDED_DRONE_MEDIA_EVENT))
+    # Insert flight/session events before the per-video events.
+    events = [project_event] + list(flight_events.values()) + events[1:]
+    out = OUT_DIR / "drones"
+    write_csv(out / "darwin_core_events_drones.csv", events, list(events[0].keys()))
+    write_csv(out / "darwin_core_occurrences_drones.csv", occurrences, OCC_FIELDS)
+    return len(events), len(occurrences)
+
+
+def copy_existing_modality(modality: str, files: list[str]) -> tuple[int, int]:
+    src = BASE_DIR / "processing_output/darwin_core" / modality
+    dst = OUT_DIR / modality
+    dst.mkdir(parents=True, exist_ok=True)
+    events_n = occ_n = 0
+    for filename in files:
+        src_file = src / filename
+        dst_file = dst / filename
+        shutil.copy2(src_file, dst_file)
+        if filename.endswith(".csv"):
+            n = len(read_csv(dst_file))
+            if "events" in filename:
+                events_n += n
+            elif "occurrences" in filename:
+                occ_n += n
+    return events_n, occ_n
+
+
+def patch_citizen_science_provenance() -> tuple[int, int]:
+    """Clarify citizen-science raw, curated, and external-album media provenance."""
+    events_path = OUT_DIR / "citizen_science/darwin_core_events_citizen_science.csv"
+    occurrences_path = OUT_DIR / "citizen_science/darwin_core_occurrences_citizen_science.csv"
+    raw_dir = BASE_DIR / "thewilds_20250804_20250828/citizen_science/photos_videos"
+    events = read_csv(events_path)
+    for row in events:
+        if row.get("eventID") == "TheWilds_2025_citizen_science":
+            row["sampleSizeValue"] = "564"
+            row["sampleSizeUnit"] = "curated wildlife photographs"
+            row["fieldNotes"] = ("Raw-folder inventory: 1,103 media files (965 still images and 138 videos). "
+                                 "Manual annotations cover 791 photos; curation retained 564 wildlife photos. "
+                                 "Multi-taxon photos yield 622 Darwin Core Occurrences across 13 observation sessions. "
+                                 "Fifty curated Sep. 8 photos are sourced from the Google Photos album export rather "
+                                 "than the photos_videos folder; their media references state that source explicitly.")
+    write_csv(events_path, events, EVENT_FIELDS)
+    occurrences = read_csv(occurrences_path)
+    external = 0
+    for row in occurrences:
+        filename = Path(row.get("associatedMedia", "")).name
+        if filename and not (raw_dir / filename).is_file():
+            row["associatedMedia"] = f"Google Photos Takeout/The Wilds 9-8-25/{filename}"
+            note = "media_source=Google Photos Takeout/The Wilds 9-8-25; not in photos_videos raw-folder inventory"
+            if note not in row.get("occurrenceRemarks", ""):
+                row["occurrenceRemarks"] = "; ".join(part for part in [row.get("occurrenceRemarks", ""), note] if part)
+            external += 1
+    if external != 50:
+        raise ValueError(f"Expected 50 externally sourced citizen-science media references, found {external}")
+    write_csv(occurrences_path, occurrences, OCC_FIELDS)
+    audit = {
+        "reconciliation": "citizen-science media populations and provenance",
+        "raw_folder": "thewilds_20250804_20250828/citizen_science/photos_videos",
+        "raw_folder_media_files": 1103,
+        "raw_folder_file_types": {"jpeg": 507, "jpg": 346, "heic": 112, "mov": 91, "mp4": 47},
+        "raw_folder_still_images": 965,
+        "raw_folder_videos": 138,
+        "manual_annotation_input_photos": 791,
+        "curated_wildlife_photos": 564,
+        "darwin_core_occurrences": 622,
+        "observation_sessions": 13,
+        "curated_media_in_raw_folder": 514,
+        "curated_media_from_google_photos_album_only": 50,
+        "external_album_media_date": "2025-09-08",
+        "external_album_media_reference_prefix": "Google Photos Takeout/The Wilds 9-8-25/",
+        "retired_unreproducible_counts": ["~991 submissions", "1,128 photographs"],
+    }
+    (OUT_DIR / "citizen_science_media_reconciliation_20261008.json").write_text(
+        json.dumps(audit, indent=2) + "\n", encoding="utf-8"
+    )
+    return len(events), len(occurrences)
+
+
+def merge_outputs() -> tuple[int, int]:
+    event_files = [
+        OUT_DIR / "camera_traps/darwin_core_events_camera_traps.csv",
+        OUT_DIR / "bioacoustics_complete/darwin_core_events_bioacoustics_complete.csv",
+        OUT_DIR / "drones/darwin_core_events_drones.csv",
+        OUT_DIR / "citizen_science/darwin_core_events_citizen_science.csv",
+        OUT_DIR / "text_data/darwin_core_events_text.csv",
+    ]
+    occ_files = [
+        ("camera_traps", OUT_DIR / "camera_traps/darwin_core_occurrences_camera_traps.csv"),
+        ("bioacoustics", OUT_DIR / "bioacoustics_complete/darwin_core_occurrences_bioacoustics_complete.csv"),
+        ("drones", OUT_DIR / "drones/darwin_core_occurrences_drones.csv"),
+        ("citizen_science", OUT_DIR / "citizen_science/darwin_core_occurrences_citizen_science.csv"),
+        ("text_data", OUT_DIR / "text_data/darwin_core_occurrences_text.csv"),
+    ]
+    events = [dict(STUDY_ROOT_EVENT)]
+    for path in event_files:
+        events.extend(read_csv(path))
+    occ = []
+    fields = ["modality_type"] + OCC_FIELDS
+    for modality, path in occ_files:
+        for row in read_csv(path):
+            out = {field: row.get(field, "") for field in OCC_FIELDS}
+            out["modality_type"] = modality
+            occ.append(out)
+    write_csv(OUT_DIR / "events_all_modalities.csv", events, EVENT_FIELDS)
+    write_csv(OUT_DIR / "occurrences_all_modalities.csv", occ, fields)
+    return len(events), len(occ)
+
+
+def valid_event_date(value: str) -> bool:
+    """Accept one ISO date or an ISO date interval, rejecting known sentinels."""
+    value = (value or "").strip()
+    if value.lower() in DATE_SENTINELS:
+        return False
+    parts = value.split("/")
+    if not parts or len(parts) > 2:
+        return False
+    try:
+        for part in parts:
+            if part.lower() in DATE_SENTINELS:
+                return False
+            date.fromisoformat(part)
+    except ValueError:
+        return False
+    return True
+
+
+def validate_snapshot() -> list[str]:
+    notes = []
+    event_ids = set()
+    for path in OUT_DIR.glob("*/darwin_core_events*.csv"):
+        rows = read_csv(path)
+        event_ids.update(row.get("eventID", "") for row in rows)
+        missing_date = sum(1 for row in rows if not row.get("eventDate"))
+        invalid_date = sum(1 for row in rows if not valid_event_date(row.get("eventDate", "")))
+        notes.append(
+            f"- `{path.relative_to(OUT_DIR)}`: rows={len(rows)}, "
+            f"missing_eventDate={missing_date}, invalid_eventDate={invalid_date}"
+        )
+    for path in OUT_DIR.glob("*/darwin_core_occurrences*.csv"):
+        rows = read_csv(path)
+        missing_event = sum(1 for row in rows if row.get("eventID") not in event_ids)
+        missing_taxon = sum(1 for row in rows if not (row.get("scientificName") or row.get("vernacularName")))
+        missing_date = sum(1 for row in rows if not row.get("eventDate"))
+        invalid_date = sum(1 for row in rows if not valid_event_date(row.get("eventDate", "")))
+        dup_ids = len(rows) - len({row.get("occurrenceID") for row in rows})
+        notes.append(
+            f"- `{path.relative_to(OUT_DIR)}`: rows={len(rows)}, "
+            f"missing_event_links={missing_event}, missing_taxon={missing_taxon}, "
+            f"missing_eventDate={missing_date}, invalid_eventDate={invalid_date}, "
+            f"duplicate_occurrenceID={dup_ids}"
+        )
+    return notes
+
+def write_qc_summary(counts: dict[str, tuple[int, int]], merged: tuple[int, int]) -> None:
+    lines = [
+        "# Darwin Core QC Summary",
+        "",
+        f"Generated: {datetime.now().isoformat(timespec='seconds')}",
+        "",
+        "## Counts",
+        "",
+    ]
+    for modality, (events, occ) in counts.items():
+        lines.append(f"- {modality}: events={events}, occurrences={occ}")
+    lines.append(f"- merged: events={merged[0]}, occurrences={merged[1]}")
+    lines.extend([
+        "",
+        "## Validation",
+        "",
+        *validate_snapshot(),
+        "",
+        "## Notes",
+        "",
+        "- Camera traps use SAM3 detections, BioCLIP Tree-of-Life labels, the current camera timestamp map, and model-derived animal geolocation estimates.",
+        "- Bioacoustic scientific names are filled from BirdNET result CSV common-name mappings.",
+        "- Drone occurrences use per-video SAM3 + BioCLIP Tree-of-Life prediction CSVs after first-pass drone ground-position projection.",
+        "- Drone decimalLatitude/decimalLongitude use projected animal coordinates only when `drone_georef_status=ok`; otherwise they fall back to frame/drone telemetry coordinates.",
+        "- Citizen-science and text-data outputs are copied from the existing Darwin Core outputs.",
+        "- Camera-trap animal coordinates are estimated positions, not direct GPS observations; see `dataGeneralizations` and geolocation provenance columns in the archived prediction CSVs.",
+    ])
+    (OUT_DIR / "QC_SUMMARY.md").write_text("\n".join(lines) + "\n", encoding="utf-8")
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--force", action="store_true", help="Overwrite existing generated snapshot outputs under OUT_DIR.")
+    args = parser.parse_args()
+
+    ensure_clean_output(args.force)
+    counts = {}
+    counts["camera_traps"] = generate_camera_traps()
+    counts["bioacoustics_complete"] = patch_bioacoustics()
+    counts["drones"] = generate_drones()
+    counts["citizen_science"] = copy_existing_modality("citizen_science", [
+        "darwin_core_events_citizen_science.csv",
+        "darwin_core_occurrences_citizen_science.csv",
+    ])
+    counts["citizen_science"] = patch_citizen_science_provenance()
+    counts["text_data"] = copy_existing_modality("text_data", [
+        "darwin_core_events_text.csv",
+        "darwin_core_occurrences_text.csv",
+        "site_selection_rationale.csv",
+    ])
+    merged = merge_outputs()
+    write_qc_summary(counts, merged)
+    print(f"Wrote Darwin Core tables to {OUT_DIR}")
+    for modality, (events, occ) in counts.items():
+        print(f"  {modality}: events={events}, occurrences={occ}")
+    print(f"  merged: events={merged[0]}, occurrences={merged[1]}")
+
+
+if __name__ == "__main__":
+    main()

smartwilds-citizen-science/CHECKSUMS.sha256

@@ -0,0 +1,6 @@
+f1091f6f362e6a7b3f78705ffc123685a7406882ba4dd97982af8478d3a0a4d2  data/events.csv
+845317eaa1fe41344a7ffe4b76a8bf23c70961818f4726bb2651e6efeb99da2e  data/occurrences.csv
+652cb6f461bfd55607183d5afa2f90ae66eb01d7c94ed3b9e1b65d1730d96a38  media_manifest.csv
+5225f2f37cd8738e7ccc72efeea6cd58389ec7d1a621d1023a481b10f49c076e  metadata/DATA_DICTIONARY.md
+7e64adcc08bb91f449f73ab5737c29c73b1500f7ad56e371bcf44b8e8ca3c6ab  metadata/provenance/citizen_science_media_reconciliation_20261008.json
+83cba7ad8d340c28ac25519c9bf7845e034dfcdd6ef16e8a62fedd3f8a55293d  scripts/build_hf_release.py

smartwilds-citizen-science/README.md

@@ -0,0 +1,161 @@
+---
+license: cc0-1.0
+language:
+- en
+pretty_name: SmartWilds Citizen Science (The Wilds 2025)
+task_categories:
+- image-classification
+tags:
+- biology
+- ecology
+- biodiversity
+- citizen-science
+- darwin-core
+- fair2
+- smartwilds
+- the-wilds
+- ohio
+size_categories:
+- n<1K
+description: >-
+  Visitor-style wildlife photographs, taken by project researchers, from 13
+  observation sessions at The
+  Wilds (Ohio, USA), 2025: 1,103 raw media files, 564 curated wildlife photos,
+  and 622 expert-verified Darwin Core Occurrences.
+
+configs:
+- config_name: occurrences
+  data_files: data/occurrences.csv
+- config_name: events
+  data_files: data/events.csv
+---
+
+# Dataset Card for SmartWilds Citizen Science
+
+![Citizen-science photos of three managed focal species](images/smartwilds-citizen-science/preview.png)
+*Visitor photographs of Przewalski's horse, Sichuan takin, and Milu (Père David's) deer, with annotation boxes.*
+
+
+This is the citizen-science dataset of the **SmartWilds** collection. Each
+modality has its own dataset and DOI. They share one Darwin Core Event hierarchy
+rooted at `TheWilds_2025` and are joined by the SmartWilds collection.
+
+## Dataset Details
+
+- **Curated by:** Jenna M. Kline, Anirudh Potlapally, Elizabeth G. Campolongo
+- **Homepage:** <https://github.com/Imageomics/smartwilds> (documentation: <https://imageomics.github.io/smartwilds/>)
+- **Paper:** SmartWilds: Integrating, Aligning, and Documenting Multi-Modal Field Sensing Data (under review)
+- **Release build:** 2026-10-22
+
+Members of the project team, moving through the park as visitors do,
+submitted photographs through a structured web form; the team holds the
+rights to the media. The release distinguishes four counts, each a different
+unit:
+
+| Unit | Count |
+|---|---:|
+| Raw media files in `photos_videos` (965 images, 138 videos) | 1,103 |
+| Photos manually annotated | 791 |
+| Curated wildlife photos | 564 |
+| Darwin Core Occurrences (some photos contain several taxa) | 622 |
+
+Fifty of the curated photos, all from September 8, came from a Google Photos
+album export that is not part of the project archive. Their 50 Occurrences are
+released, but the photos are not: `associatedMedia` is blank for these rows,
+and `occurrenceRemarks` gives the original album path and the reason
+(`media_not_distributed`).
+
+The Event table has 14 rows: 1 `Project` and 13 `ObservationSession`.
+
+### Annotation method
+
+- **Method:** manual. All 622 Occurrences are `verified by expert`
+  (`HumanObservation`).
+- **Locality:** coordinates are generalized to locality level.
+
+### Personal and sensitive information
+
+The photographs were taken by project researchers, not by members of the
+public, so no third-party consent is needed for their release. Some show
+other visitors. Media in which a person can be individually identified are removed from the
+release. People too small to identify (a person box under 80 px in the native
+image, where the face is under about 10 px) are retained; every file with a
+larger person detection was reviewed by a person before release.
+
+### Supported Tasks and Applications
+
+This dataset supports ecological analysis and sensing-systems research; the computer-vision tasks apply once the photographs are released.
+
+**🌿 Ecological Applications:**
+- Expert-verified species records (622) as a validation set for the machine-labeled modalities
+- Visitor-perspective sampling bias: which taxa and places opportunistic photography captures
+- Cross-modal corroboration of focal-species presence by date
+
+**⚙️ Sensing-Systems Research:**
+- Sparse, bursty human-reported arrivals as priors for sensor scheduling (citizen-science-as-prior case study)
+- Session-level effort and timing from observation sessions
+
+**🤖 Computer Vision Tasks (with the photo release):**
+- Species classification on smartphone imagery with verified labels
+- Detection on photographs with annotation boxes
+
+## Considerations for Using the Data
+
+- **Submissions arrive only during operating hours.** The diurnal profile is
+  truncated by visitor schedules, not by animal activity.
+- **Visitors' attention skews the taxa.** Charismatic and easily seen taxa are
+  over-represented.
+
+## Supplemental Material
+
+Extended inventories, the timestamp-repair methods, the standards crosswalk,
+validation checks, and the full defect inventory are in
+[`SUPPLEMENTAL.md`](https://huggingface.co/datasets/ABC-Center/smartwilds_metadata/blob/main/SUPPLEMENTAL.md)
+in the study-level SmartWilds dataset.
+
+## Licensing
+
+Data: [CC0 1.0](https://creativecommons.org/publicdomain/zero/1.0/). Scripts: MIT.
+
+## Citation
+
+```bibtex
+@misc{smartwilds2025_citizen_science,
+  author    = {Kline, Jenna M. and Potlapally, Anirudh and Campolongo, Elizabeth G.},
+  title     = {SmartWilds Citizen Science (The Wilds 2025)},
+  year      = {2026},
+  publisher = {Hugging Face},
+  note      = {DOI pending the raw data release},
+  url       = {https://huggingface.co/datasets/ABC-Center/smartwilds_metadata/tree/main/smartwilds-citizen-science}
+}
+```
+
+Please also cite the paper:
+
+```bibtex
+@article{kline2026smartwilds,
+  author  = {Kline, Jenna M. and Potlapally, Anirudh and Campolongo, Elizabeth G.},
+  title   = {SmartWilds: Integrating, Aligning, and Documenting Multi-Modal Field Sensing Data},
+  year    = {2026},
+  note    = {Under review}
+}
+```
+
+This dataset card follows the [FAIR² Drones Data Standard](https://imageomics.github.io/fair_drones/)
+(Kline and Campolongo, 2026).
+
+## Acknowledgements
+
+This project is supported by the [Imageomics Institute](https://imageomics.org), which is funded by the US National Science Foundation's Harnessing the Data Revolution (HDR) program under [Award #2118240](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2118240) (Imageomics: A New Frontier of Biological Information Powered by Knowledge-Guided Machine Learning), and by the [AI Institute for Intelligent Cyberinfrastructure with Computational Learning in the Environment (ICICLE)](https://icicle.osu.edu/), which is funded by the US National Science Foundation under [Award #2112606](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2112606).
+
+This work is also supported by the AI and Biodiversity Change (ABC) Global Center. The ABC Global Center is funded by the US National Science Foundation under [Award No. 2330423](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2330423&HistoricalAwards=false) and the Natural Sciences and Engineering Research Council of Canada under [Award No. 585136](https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=782440). This dataset draws on research supported by the Social Sciences and Humanities Research Council.
+
+Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation, Natural Sciences and Engineering Research Council of Canada, or Social Sciences and Humanities Research Council.
+
+Special thanks to Dan Beetem and [The Wilds](https://www.thewilds.org/) in Ohio for providing access to deployment sites and supporting this wildlife monitoring research. We thank the Columbus Zoo and Aquarium for their support in facilitating this project. All data collection, including drone flights, was conducted under the supervision of the Director of Animal Management, with permission from The Wilds Animal Care and Use Committee.
+
+This work used the [Ohio Supercomputer Center](https://www.osc.edu/) (Ohio Supercomputer Center. 1987. Ohio Supercomputer Center. Columbus OH: Ohio Supercomputer Center. <https://ror.org/01apna436>).
+
+## Dataset Card Contact
+
+Jenna M. Kline (kline.377@osu.edu)

smartwilds-citizen-science/data/events.csv

@@ -0,0 +1,15 @@
+eventID,parentEventID,eventType,eventDate,samplingProtocol,sampleSizeValue,sampleSizeUnit,locality,decimalLatitude,decimalLongitude,coordinateUncertaintyInMeters,geodeticDatum,country,stateProvince,county,datasetID,datasetName,recordedBy,fieldNotes,eventTime
+TheWilds_2025_citizen_science,TheWilds_2025,Project,2025-06-30/2025-09-08,Opportunistic photography by visiting researchers and students,564,curated wildlife photographs,"The Wilds Conservation Center, Muskingum County, Ohio, USA",39.83,-81.74,5000,WGS84,United States,Ohio,Muskingum,TheWilds_2025_citizen_science,The Wilds 2025 Multi-Modal Survey - Citizen Science,Multiple citizen observers,"Raw-folder inventory: 1,103 media files (965 still images and 138 videos). Manual annotations cover 791 photos; curation retained 564 wildlife photos (excluded 137 with humans, 77 equipment-only, and 225 without species; categories overlap). Multi-taxon photos yield 622 Darwin Core Occurrences across 13 observation sessions. Fifty curated Sep. 8 photos are sourced from the Google Photos album export rather than the photos_videos folder; their media references state that source explicitly. 20 species identified via manual annotation.",
+TheWilds_2025_citizen_science_2025-06-30,TheWilds_2025_citizen_science,ObservationSession,2025-06-30,Opportunistic photography,9,photographs,The Wilds Conservation Center,39.83,-81.74,5000,WGS84,United States,Ohio,Muskingum,TheWilds_2025_citizen_science,The Wilds 2025 Multi-Modal Survey - Citizen Science,Citizen observers,Campaign: 20250630_20250703. 9 photos on 2025-06-30.,11:28:20/14:21:47
+TheWilds_2025_citizen_science_2025-07-01,TheWilds_2025_citizen_science,ObservationSession,2025-07-01,Opportunistic photography,301,photographs,The Wilds Conservation Center,39.83,-81.74,5000,WGS84,United States,Ohio,Muskingum,TheWilds_2025_citizen_science,The Wilds 2025 Multi-Modal Survey - Citizen Science,Citizen observers,Campaign: 20250630_20250703. 301 photos on 2025-07-01.,02:34:34/10:09:35
+TheWilds_2025_citizen_science_2025-08-04,TheWilds_2025_citizen_science,ObservationSession,2025-08-04,Opportunistic photography,12,photographs,The Wilds Conservation Center,39.83,-81.74,5000,WGS84,United States,Ohio,Muskingum,TheWilds_2025_citizen_science,The Wilds 2025 Multi-Modal Survey - Citizen Science,Citizen observers,Campaign: 20250804_20250828. 12 photos on 2025-08-04.,09:44:10/09:45:53
+TheWilds_2025_citizen_science_2025-08-07,TheWilds_2025_citizen_science,ObservationSession,2025-08-07,Opportunistic photography,28,photographs,The Wilds Conservation Center,39.83,-81.74,5000,WGS84,United States,Ohio,Muskingum,TheWilds_2025_citizen_science,The Wilds 2025 Multi-Modal Survey - Citizen Science,Citizen observers,Campaign: 20250804_20250828. 28 photos on 2025-08-07.,10:00:00/23:16:31
+TheWilds_2025_citizen_science_2025-08-09,TheWilds_2025_citizen_science,ObservationSession,2025-08-09,Opportunistic photography,1,photographs,The Wilds Conservation Center,39.83,-81.74,5000,WGS84,United States,Ohio,Muskingum,TheWilds_2025_citizen_science,The Wilds 2025 Multi-Modal Survey - Citizen Science,Citizen observers,Campaign: 20250804_20250828. 1 photos on 2025-08-09.,13:43:59
+TheWilds_2025_citizen_science_2025-08-11,TheWilds_2025_citizen_science,ObservationSession,2025-08-11,Opportunistic photography,1,photographs,The Wilds Conservation Center,39.83,-81.74,5000,WGS84,United States,Ohio,Muskingum,TheWilds_2025_citizen_science,The Wilds 2025 Multi-Modal Survey - Citizen Science,Citizen observers,Campaign: 20250804_20250828. 1 photos on 2025-08-11.,09:38:44
+TheWilds_2025_citizen_science_2025-08-14,TheWilds_2025_citizen_science,ObservationSession,2025-08-14,Opportunistic photography,1,photographs,The Wilds Conservation Center,39.83,-81.74,5000,WGS84,United States,Ohio,Muskingum,TheWilds_2025_citizen_science,The Wilds 2025 Multi-Modal Survey - Citizen Science,Citizen observers,Campaign: 20250804_20250828. 1 photos on 2025-08-14.,09:27:57
+TheWilds_2025_citizen_science_2025-08-18,TheWilds_2025_citizen_science,ObservationSession,2025-08-18,Opportunistic photography,8,photographs,The Wilds Conservation Center,39.83,-81.74,5000,WGS84,United States,Ohio,Muskingum,TheWilds_2025_citizen_science,The Wilds 2025 Multi-Modal Survey - Citizen Science,Citizen observers,Campaign: 20250804_20250828. 8 photos on 2025-08-18.,09:37:09/09:41:07
+TheWilds_2025_citizen_science_2025-08-21,TheWilds_2025_citizen_science,ObservationSession,2025-08-21,Opportunistic photography,91,photographs,The Wilds Conservation Center,39.83,-81.74,5000,WGS84,United States,Ohio,Muskingum,TheWilds_2025_citizen_science,The Wilds 2025 Multi-Modal Survey - Citizen Science,Citizen observers,Campaign: 20250804_20250828. 91 photos on 2025-08-21.,09:23:01/12:14:38
+TheWilds_2025_citizen_science_2025-08-22,TheWilds_2025_citizen_science,ObservationSession,2025-08-22,Opportunistic photography,14,photographs,The Wilds Conservation Center,39.83,-81.74,5000,WGS84,United States,Ohio,Muskingum,TheWilds_2025_citizen_science,The Wilds 2025 Multi-Modal Survey - Citizen Science,Citizen observers,Campaign: 20250804_20250828. 14 photos on 2025-08-22.,05:33:27/09:53:25
+TheWilds_2025_citizen_science_2025-08-26,TheWilds_2025_citizen_science,ObservationSession,2025-08-26,Opportunistic photography,8,photographs,The Wilds Conservation Center,39.83,-81.74,5000,WGS84,United States,Ohio,Muskingum,TheWilds_2025_citizen_science,The Wilds 2025 Multi-Modal Survey - Citizen Science,Citizen observers,Campaign: 20250804_20250828. 8 photos on 2025-08-26.,13:38:00/13:38:03
+TheWilds_2025_citizen_science_2025-08-28,TheWilds_2025_citizen_science,ObservationSession,2025-08-28,Opportunistic photography,40,photographs,The Wilds Conservation Center,39.83,-81.74,5000,WGS84,United States,Ohio,Muskingum,TheWilds_2025_citizen_science,The Wilds 2025 Multi-Modal Survey - Citizen Science,Citizen observers,Campaign: 20250804_20250828. 40 photos on 2025-08-28.,22:47:01/22:47:19
+TheWilds_2025_citizen_science_2025-09-08,TheWilds_2025_citizen_science,ObservationSession,2025-09-08,Opportunistic photography,50,photographs,The Wilds Conservation Center,39.83,-81.74,5000,WGS84,United States,Ohio,Muskingum,TheWilds_2025_citizen_science,The Wilds 2025 Multi-Modal Survey - Citizen Science,Citizen observers,Campaign: 20250908_20250908. 50 photos on 2025-09-08.,09:46:36/22:07:46

smartwilds-citizen-science/metadata/DATA_DICTIONARY.md

@@ -0,0 +1,86 @@
+# smartwilds-citizen-science data dictionary
+
+Generated by `scripts/build_hf_release.py`. Darwin Core term definitions: <https://dwc.tdwg.org/terms/>. Columns not listed in a Darwin Core namespace are SmartWilds-specific.
+
+## `data/events.csv`
+
+14 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `eventID` | Darwin Core: unique Event identifier | TheWilds_2025_citizen_science |
+| `parentEventID` | Darwin Core: eventID of the parent Event | TheWilds_2025 |
+| `eventType` | Darwin Core: Project, SensorDeployment, AudioRecording, MediaEvent, or ObservationSession | Project |
+| `eventDate` | Darwin Core: ISO 8601 date or date interval (start/end) | 2025-06-30/2025-09-08 |
+| `samplingProtocol` | Darwin Core: sensor, schedule, and processing chain | Opportunistic photography by visiting researchers and studen… |
+| `sampleSizeValue` | Darwin Core: sample size | 564 |
+| `sampleSizeUnit` | Darwin Core: unit of sampleSizeValue | curated wildlife photographs |
+| `locality` | Darwin Core: named place | The Wilds Conservation Center, Muskingum County, Ohio, USA |
+| `decimalLatitude` | Darwin Core: WGS84; see dataGeneralizations for what the point represents | 39.83 |
+| `decimalLongitude` | Darwin Core: WGS84 | -81.74 |
+| `coordinateUncertaintyInMeters` | Darwin Core: meters | 5000 |
+| `geodeticDatum` | Darwin Core: WGS84 | WGS84 |
+| `country` |  | United States |
+| `stateProvince` |  | Ohio |
+| `county` |  | Muskingum |
+| `datasetID` | Darwin Core: modality dataset identifier | TheWilds_2025_citizen_science |
+| `datasetName` | Darwin Core | The Wilds 2025 Multi-Modal Survey - Citizen Science |
+| `recordedBy` | Darwin Core: device or team | Multiple citizen observers |
+| `fieldNotes` | Darwin Core: provenance and field notes | Raw-folder inventory: 1,103 media files (965 still images an… |
+| `eventTime` | Darwin Core: local time (or interval); blank when not recoverable | 11:28:20/14:21:47 |
+
+## `data/occurrences.csv`
+
+622 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `occurrenceID` | Darwin Core: unique Occurrence identifier | TheWilds_2025_CS_0001 |
+| `eventID` | Darwin Core: unique Event identifier | TheWilds_2025_citizen_science_2025-08-28 |
+| `catalogNumber` | Darwin Core: source record identifier | CS_012A0E1A-022A-45E8-9FA2-BE0838AAD135_4_5005_c.jpeg_0 |
+| `scientificName` | Darwin Core: top-1 model or annotator taxon | Cervus hanglu bactrianus Lydekker, 1900 |
+| `kingdom` | Darwin Core: rank | Animalia |
+| `phylum` | Darwin Core: rank | Chordata |
+| `class` | Darwin Core: rank | Mammalia |
+| `order` | Darwin Core: rank | Artiodactyla |
+| `family` | Darwin Core: rank | Cervidae |
+| `genus` | Darwin Core: rank | Cervus |
+| `specificEpithet` | Darwin Core: rank | hanglu |
+| `taxonRank` | Darwin Core: rank of scientificName | subspecies |
+| `vernacularName` | Darwin Core: common name | Bactrian Deer |
+| `occurrenceStatus` | Darwin Core: present | present |
+| `individualCount` | Darwin Core: 1 per visual crop; number of BirdNET detections for acoustic rows | 1 |
+| `basisOfRecord` | Darwin Core: MachineObservation or HumanObservation | HumanObservation |
+| `recordedBy` | Darwin Core: device or team | Citizen observer (manual annotation) |
+| `identifiedBy` | Darwin Core: model (and version) or annotator | Jenna Kline |
+| `identificationVerificationStatus` | Darwin Core: verification status; the vocabulary differs by table (SUPPLEMENTAL.md S9.3) | verified by expert |
+| `identificationRemarks` | Darwin Core: rank, top-1 score, bbox xywh (px), detector label | Manual species identification from citizen science photograp… |
+| `eventDate` | Darwin Core: ISO 8601 date or date interval (start/end) | 2025-08-28 |
+| `eventTime` | Darwin Core: local time (or interval); blank when not recoverable | 22:47:01 |
+| `decimalLatitude` | Darwin Core: WGS84; see dataGeneralizations for what the point represents | 39.83 |
+| `decimalLongitude` | Darwin Core: WGS84 | -81.74 |
+| `coordinateUncertaintyInMeters` | Darwin Core: meters | 5000 |
+| `geodeticDatum` | Darwin Core: WGS84 | WGS84 |
+| `locality` | Darwin Core: named place | The Wilds Conservation Center, Muskingum County, Ohio, USA |
+| `country` |  | United States |
+| `stateProvince` |  | Ohio |
+| `county` |  | Muskingum |
+| `establishmentMeans` |  | managed |
+| `degreeOfEstablishment` |  | captive |
+| `occurrenceRemarks` | Darwin Core: projection status, timestamp method, repair notes (key=value; ...) | Photographed during guided visit to The Wilds. |
+| `associatedMedia` | Darwin Core: release-relative path to the source media file | media/thewilds_20250804_20250828/photos_videos/012A0E1A-022A… |
+| `datasetID` | Darwin Core: modality dataset identifier | TheWilds_2025_citizen_science |
+| `datasetName` | Darwin Core | The Wilds 2025 Multi-Modal Survey - Citizen Science |
+
+## `media_manifest.csv`
+
+1,104 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `release_path` | Path of the file inside this dataset | media/thewilds_20250804_20250828/heic_to_jpeg_conversion_man… |
+| `source_path` | Path in the project archive | thewilds_20250804_20250828/citizen_science/heic_to_jpeg_conv… |
+| `bytes` | File size | 26259 |
+| `include` | true if the file is distributed in media/ | true |
+| `exclusion_reason` | Why a file is not distributed |  |
+

smartwilds-citizen-science/metadata/provenance/citizen_science_media_reconciliation_20261008.json

@@ -0,0 +1,26 @@
+{
+  "reconciliation": "citizen-science media populations and provenance",
+  "raw_folder": "thewilds_20250804_20250828/citizen_science/photos_videos",
+  "raw_folder_media_files": 1103,
+  "raw_folder_file_types": {
+    "jpeg": 507,
+    "jpg": 346,
+    "heic": 112,
+    "mov": 91,
+    "mp4": 47
+  },
+  "raw_folder_still_images": 965,
+  "raw_folder_videos": 138,
+  "manual_annotation_input_photos": 791,
+  "curated_wildlife_photos": 564,
+  "darwin_core_occurrences": 622,
+  "observation_sessions": 13,
+  "curated_media_in_raw_folder": 514,
+  "curated_media_from_google_photos_album_only": 50,
+  "external_album_media_date": "2025-09-08",
+  "external_album_media_reference_prefix": "Google Photos Takeout/The Wilds 9-8-25/",
+  "retired_unreproducible_counts": [
+    "~991 submissions",
+    "1,128 photographs"
+  ]
+}

smartwilds-citizen-science/scripts/build_hf_release.py

@@ -0,0 +1,716 @@
+#!/usr/bin/env python3
+"""Stage the SmartWilds per-modality Hugging Face dataset directories.
+
+One directory per modality (camera traps, bioacoustics, drones, citizen
+science) plus the umbrella ``smartwilds`` dataset, each laid out like
+imageomics/kabr-behavior-telemetry:
+
+    <repo>/README.md                 dataset card (filled in separately)
+    <repo>/data/events.csv           Darwin Core Event table
+    <repo>/data/occurrences.csv      Darwin Core Occurrence table
+    <repo>/data/...                  modality-specific per-file tables
+    <repo>/metadata/                 sensor settings, data dictionary, provenance
+    <repo>/scripts/                  generator / georeferencing / this script
+    <repo>/media_manifest.csv        one row per raw file: release path, bytes,
+                                     include flag, exclusion reason
+
+Inputs are the frozen Darwin Core tables of the integration pass
+(``SMARTWILDS_DWC_SNAPSHOT``) and the validated post-triage camera derivative
+(``SMARTWILDS_CAMERA_SOURCE``). Neither is modified. The staged copies get one
+release repair, audited in ``RELEASE_BUILD_AUDIT.json``: ``associatedMedia`` is
+rewritten from absolute archive paths to release-relative ``media/...`` paths,
+and drone frames in the purged scratch workspace fall back to their retained
+original video.
+
+Media files are NOT copied here. ``media_manifest.csv`` is the packaging input.
+
+Locations are read from environment variables, all optional:
+
+    SMARTWILDS_ARCHIVE         field archive root (default: this repository)
+    SMARTWILDS_DWC_SNAPSHOT    frozen Darwin Core tables
+    SMARTWILDS_CAMERA_SOURCE   validated camera derivative
+    SMARTWILDS_RELEASE         output directory (also ``--out``)
+    SMARTWILDS_PUBLIC_SCRIPTS  the public scripts shipped inside each dataset
+    SMARTWILDS_SCRATCH_PREFIX  prefix of purged scratch frame paths in the tables
+    SMARTWILDS_DETECTIONS      prefix of the external detection workspace
+
+Usage:
+    python release/scripts/build_hf_release.py [--out DIR] [--only REPO ...]
+"""
+import argparse
+import hashlib
+import json
+import os
+import re
+import shutil
+import subprocess
+import sys
+from datetime import datetime, timezone
+from pathlib import Path
+
+import pandas as pd
+
+def env_path(name, default):
+    return Path(os.environ.get(name, default)).resolve()
+
+
+ROOT = env_path("SMARTWILDS_ARCHIVE", Path(__file__).resolve().parents[2])
+SNAP = env_path("SMARTWILDS_DWC_SNAPSHOT", ROOT / "processing_output/darwin_core_latest_20261008")
+CAMERA_SOURCE = env_path("SMARTWILDS_CAMERA_SOURCE", ROOT / "archive/camera_traps_validated_20261108")
+RELEASE_DIR = env_path("SMARTWILDS_RELEASE", ROOT / "release/hf_smartwilds_20261022")
+PUBLIC_SCRIPTS = env_path("SMARTWILDS_PUBLIC_SCRIPTS", ROOT / "public/smartwilds/scripts")
+TS_MAP = ROOT / "processing_output/detections/drone_camera_analysis/camera_time_mapping"
+SENSOR_MD = ROOT / "documentation/sensor_configurations.md"
+CARDS = ROOT / "release/cards"
+CAMPAIGNS = ["thewilds_20250630_20250703", "thewilds_20250804_20250828", "thewilds_20250908_20250908"]
+PROJECT_PREFIX = str(ROOT) + "/"
+# Prefixes of absolute paths that may still appear inside the frozen tables.
+SCRATCH_PREFIX = os.environ.get("SMARTWILDS_SCRATCH_PREFIX", "")      # purged drone frame workspace
+DETECTIONS_PREFIX = os.environ.get("SMARTWILDS_DETECTIONS", "")        # external detection workspaces
+if DETECTIONS_PREFIX and not DETECTIONS_PREFIX.endswith("/"):
+    DETECTIONS_PREFIX += "/"
+# Anything under the same filesystem roots as the known prefixes counts as an
+# archive path (e.g. "/fs/ess/" for "/fs/ess/<project>/TheWilds2025/").
+CLUSTER_MARKERS = sorted({"/".join(m.split("/")[:3]) + "/" for m in (PROJECT_PREFIX, SCRATCH_PREFIX, DETECTIONS_PREFIX)
+                          if m and m.startswith("/")})
+STUDY_START, STUDY_END = "2025-06-30", "2025-09-08"
+
+REPOS = {
+    "camera_traps": "smartwilds-camera-traps",
+    "bioacoustics": "smartwilds-bioacoustics",
+    "drones": "smartwilds-drones",
+    "citizen_science": "smartwilds-citizen-science",
+    "core": "smartwilds",
+}
+
+audit = {"inputs": {"snapshot": str(SNAP), "camera_source": str(CAMERA_SOURCE)}, "built_utc": None,
+         "repos": {}, "blocked": [], "errors": []}
+
+
+def read(path, **kw):
+    return pd.read_csv(path, dtype=str, keep_default_na=False, **kw)
+
+
+def readable(path):
+    return os.access(path, os.R_OK)
+
+
+def sha256(path, buf=1 << 20):
+    h = hashlib.sha256()
+    with open(path, "rb") as f:
+        for chunk in iter(lambda: f.read(buf), b""):
+            h.update(chunk)
+    return h.hexdigest()
+
+
+def rel_media(abs_path):
+    """<archive>/<campaign>/<modality>/<rest> -> media/<campaign>/<rest>."""
+    if not abs_path.startswith(PROJECT_PREFIX):
+        return None
+    parts = abs_path[len(PROJECT_PREFIX):].split("/")
+    if parts[0] not in CAMPAIGNS or len(parts) < 3:
+        return None
+    return "/".join(["media", parts[0]] + parts[2:])
+
+
+BIRD_WATER = ROOT / "processing_output/detections/drone_camera_analysis/aquatic_bird_analysis/bird_water_association_per_box.csv"
+PATH_REWRITES = [(str(BIRD_WATER), "analysis/bird_water_association_per_box.csv")]
+if DETECTIONS_PREFIX:
+    PATH_REWRITES.append((DETECTIONS_PREFIX, "external_detections/"))
+PATH_REWRITES.append((PROJECT_PREFIX, "TheWilds2025/"))
+
+
+def has_cluster_path(col):
+    return any(col.str.contains(m, regex=False).any() for m in CLUSTER_MARKERS)
+
+
+def count_cluster_paths(df):
+    return int(sum(df[c].str.contains(m, regex=False).sum() for c in df.columns for m in CLUSTER_MARKERS))
+
+
+def strip_cluster_paths(df):
+    """Rewrite remaining absolute archive paths in free-text columns to
+    release-relative (shipped) or project-relative (provenance-only) forms."""
+    for c in df.columns:
+        col = df[c]
+        if has_cluster_path(col):
+            for a, b in PATH_REWRITES:
+                col = col.str.replace(a, b, regex=False)
+            df[c] = col
+    return df
+
+
+def scrub_text_copy(src, dst, extra=()):
+    """Copy a text file (JSON, Markdown), rewriting archive paths and the
+    provenance tokens of the camera derivative."""
+    text = Path(src).read_text()
+    rewrites = list(PATH_REWRITES) + list(ARRIVAL_TOKEN_REWRITES) + [(SNAP.name, "integration_pass_20261022")] + list(extra)
+    for a, b in rewrites:
+        text = text.replace(a, b)
+    Path(dst).write_text(text)
+
+
+# Files written by earlier builds that the current build no longer produces.
+STALE = ["scripts/generate_darwin_core_latest_20261008.py", "scripts/refresh_icde_camera_arrival_metadata.py",
+         "analysis/arrival_traces_v2/DEPRECATED_LEGACY_PRIMITIVES.md",
+         "analysis/arrival_traces_v2/UNAVAILABLE_CAMERA_IMAGE_CAPTURE_ACQUISITION.md",
+         "analysis/bird_water_association_per_box.csv"]
+
+
+def atomic_write(path, writer):
+    """Write through a temp file and rename, so files owned by another team
+    member (and read-only to us) are replaced rather than opened for writing."""
+    path.parent.mkdir(parents=True, exist_ok=True)
+    temp = path.with_name(f".{path.name}.tmp")
+    try:
+        writer(temp)
+        temp.replace(path)
+    finally:
+        temp.unlink(missing_ok=True)
+
+
+def parse_md_tables(md_path):
+    """Return {"## section / ### heading": DataFrame} for every pipe table."""
+    tables, section, heading, rows = {}, "", None, []
+    for line in Path(md_path).read_text().splitlines() + [""]:
+        if line.startswith("## "):
+            section = heading = line[3:].strip()
+        elif line.startswith("### "):
+            heading = section + " / " + line[4:].strip()
+        if line.startswith("|"):
+            cells = [c.strip() for c in line.strip().strip("|").split("|")]
+            if not all(re.fullmatch(r":?-+:?", c) for c in cells):
+                rows.append(cells)
+        elif rows:
+            tables[heading] = pd.DataFrame(rows[1:], columns=rows[0])
+            rows = []
+    return tables
+
+
+def media_inventory(modality, exclude=lambda rel: None):
+    """One row per raw file under every campaign's <modality>/ directory."""
+    rows = []
+    for camp in CAMPAIGNS:
+        base = ROOT / camp / modality
+        if not base.is_dir():
+            continue
+        for dirpath, _, files in os.walk(base):
+            for fn in files:
+                src = Path(dirpath) / fn
+                rel = rel_media(str(src))
+                reason = exclude(rel) or ""
+                try:
+                    size = src.stat().st_size
+                except OSError:
+                    size, reason = "", reason or "unreadable_source"
+                rows.append({"release_path": rel, "source_path": str(src.relative_to(ROOT)),
+                             "bytes": size, "include": "false" if reason else "true",
+                             "exclusion_reason": reason})
+    return pd.DataFrame(rows).sort_values("release_path").reset_index(drop=True)
+
+
+def check_tables(name, ev, oc, manifest, extra_event_ids=()):
+    """Release checks the snapshot audit lacked: dates inside parent spans,
+    sentinel strings, and every associatedMedia resolving to an included file."""
+    res = {}
+    ids = set(ev.eventID) | set(extra_event_ids)
+    res["events"], res["occurrences"] = len(ev), len(oc)
+    res["dangling_parentEventID"] = sorted({p for p in ev.parentEventID if p and p not in ids})
+    res["occurrences_without_event"] = int((~oc.eventID.isin(ids)).sum())
+    sentinel = re.compile(r"^(nan|none|null|-1|1970-01-01)$", re.I)
+    res["sentinel_cells"] = {f"{t}.{c}": int(df[c].str.match(sentinel).sum())
+                             for t, df in (("events", ev), ("occurrences", oc)) for c in df.columns
+                             if df[c].str.match(sentinel).any()}
+    span = ev.set_index("eventID").eventDate.str.split("/", expand=True)
+    span.columns = ["a", "b"][: span.shape[1]]
+    if "b" not in span:
+        span["b"] = None
+    span["b"] = span.b.fillna(span.a)
+    single = oc[~oc.eventDate.str.contains("/")]
+    j = single.join(span, on="eventID")
+    res["occurrence_date_outside_event_span"] = int(((j.eventDate < j.a) | (j.eventDate > j.b)).sum())
+    res["occurrence_date_outside_study"] = int(((single.eventDate < STUDY_START) | (single.eventDate > STUDY_END)).sum())
+    res["occurrence_interval_dates"] = int(oc.eventDate.str.contains("/").sum())
+    media = oc.associatedMedia[oc.associatedMedia != ""]
+    res["absolute_media_paths"] = int(media.str.startswith("/").sum())
+    res["cells_with_cluster_paths"] = count_cluster_paths(ev) + count_cluster_paths(oc)
+    if manifest is not None:
+        included = set(manifest.release_path[manifest.include == "true"])
+        unresolved = media[~media.isin(included)]
+        res["media_paths_not_in_included_manifest"] = int(len(unresolved))
+        res["media_unresolved_examples"] = unresolved.drop_duplicates().head(5).tolist()
+    res["pass"] = (not res["dangling_parentEventID"] and res["occurrences_without_event"] == 0
+                   and not res["sentinel_cells"] and res["occurrence_date_outside_event_span"] == 0
+                   and res["occurrence_date_outside_study"] == 0 and res["absolute_media_paths"] == 0
+                   and res["cells_with_cluster_paths"] == 0
+                   and res.get("media_paths_not_in_included_manifest", 0) == 0)
+    return res
+
+
+def write_repo(out, repo, files):
+    d = out / repo
+    for rel, obj in files.items():
+        p = d / rel
+        p.parent.mkdir(parents=True, exist_ok=True)
+        if isinstance(obj, pd.DataFrame):
+            atomic_write(p, lambda t: obj.to_csv(t, index=False))
+        elif isinstance(obj, (dict, list)):
+            atomic_write(p, lambda t: t.write_text(json.dumps(obj, indent=2) + "\n"))
+        elif isinstance(obj, Path) and obj.suffix in (".json", ".md"):
+            atomic_write(p, lambda t: scrub_text_copy(obj, t))   # provenance records: no archive paths or snapshot names
+        elif isinstance(obj, Path):
+            atomic_write(p, lambda t: shutil.copyfile(obj, t))
+        else:
+            atomic_write(p, lambda t: t.write_text(obj))
+    return d
+
+
+TERMS = {
+    # Darwin Core / Humboldt terms (https://dwc.tdwg.org/terms/)
+    "eventID": "Darwin Core: unique Event identifier", "parentEventID": "Darwin Core: eventID of the parent Event",
+    "eventType": "Darwin Core: Project, SensorDeployment, AudioRecording, MediaEvent, or ObservationSession",
+    "eventDate": "Darwin Core: ISO 8601 date or date interval (start/end)",
+    "eventTime": "Darwin Core: local time (or interval); blank when not recoverable",
+    "samplingProtocol": "Darwin Core: sensor, schedule, and processing chain", "sampleSizeValue": "Darwin Core: sample size",
+    "sampleSizeUnit": "Darwin Core: unit of sampleSizeValue", "fieldNotes": "Darwin Core: provenance and field notes",
+    "occurrenceID": "Darwin Core: unique Occurrence identifier", "catalogNumber": "Darwin Core: source record identifier",
+    "scientificName": "Darwin Core: top-1 model or annotator taxon", "taxonRank": "Darwin Core: rank of scientificName",
+    "vernacularName": "Darwin Core: common name", "kingdom": "Darwin Core: rank", "phylum": "Darwin Core: rank",
+    "class": "Darwin Core: rank", "order": "Darwin Core: rank", "family": "Darwin Core: rank", "genus": "Darwin Core: rank",
+    "specificEpithet": "Darwin Core: rank", "occurrenceStatus": "Darwin Core: present",
+    "individualCount": "Darwin Core: 1 per visual crop; number of BirdNET detections for acoustic rows",
+    "basisOfRecord": "Darwin Core: MachineObservation or HumanObservation",
+    "recordedBy": "Darwin Core: device or team", "identifiedBy": "Darwin Core: model (and version) or annotator",
+    "identificationVerificationStatus": "Darwin Core: verification status; the vocabulary differs by table (SUPPLEMENTAL.md S9.3)",
+    "identificationRemarks": "Darwin Core: rank, top-1 score, bbox xywh (px), detector label",
+    "decimalLatitude": "Darwin Core: WGS84; see dataGeneralizations for what the point represents",
+    "decimalLongitude": "Darwin Core: WGS84", "coordinateUncertaintyInMeters": "Darwin Core: meters",
+    "geodeticDatum": "Darwin Core: WGS84", "locality": "Darwin Core: named place",
+    "locationID": "Darwin Core: site (TWxx) for fixed sensors; flight ID for drones",
+    "occurrenceRemarks": "Darwin Core: projection status, timestamp method, repair notes (key=value; ...)",
+    "associatedMedia": "Darwin Core: release-relative path to the source media file",
+    "associatedReferences": "Darwin Core: release-relative supporting analysis file",
+    "dataGeneralizations": "Darwin Core: what the coordinate represents", "informationWithheld": "Darwin Core",
+    "datasetID": "Darwin Core: modality dataset identifier", "datasetName": "Darwin Core",
+    "release_path": "Path of the file inside this dataset", "source_path": "Path in the project archive",
+    "bytes": "File size", "include": "true if the file is distributed in media/",
+    "exclusion_reason": "Why a file is not distributed",
+    "has_human_detection": "true if any Occurrence on this image is labelled Homo sapiens",
+    "occurrence_count": "Occurrences derived from this image",
+    "video_path": "Release-relative path of the raw video", "detection_id": "<video>_<frame>_<k> from the SAM3 run",
+    "frame_index": "0-based native frame index in the video (SAM3 sampled every 15th frame)",
+    "bbox_x": "Box left, pixels, in the native frame", "bbox_y": "Box top, pixels", "bbox_w": "Box width, pixels",
+    "bbox_h": "Box height, pixels", "top1_score": "BioCLIP Tree-of-Life top-1 score",
+    "georef_status": "Ground-projection status (ok | invalid_altitude | missing_drone_gps | horizon_ray | out_of_range)",
+}
+
+
+def write_dictionary(d, repo):
+    out = [f"# {repo} data dictionary", "",
+           "Generated by `scripts/build_hf_release.py`. Darwin Core term definitions: "
+           "<https://dwc.tdwg.org/terms/>. Columns not listed in a Darwin Core namespace are "
+           "SmartWilds-specific.", ""]
+    for p in sorted(d.rglob("*.csv")):
+        rel = p.relative_to(d)
+        df = pd.read_csv(p, dtype=str, keep_default_na=False, nrows=1000)
+        n = sum(1 for _ in open(p)) - 1
+        out += [f"## `{rel}`", "", f"{n:,} rows.", "", "| Column | Description | Example |", "|---|---|---|"]
+        for c in df.columns:
+            ex = next((v for v in df[c] if v), "")
+            ex = (ex[:60] + "…") if len(ex) > 60 else ex
+            out.append(f"| `{c}` | {TERMS.get(c, '')} | {ex.replace('|', '/')} |")
+        out.append("")
+    atomic_write(d / "metadata/DATA_DICTIONARY.md", lambda t: t.write_text("\n".join(out) + "\n"))
+
+
+def write_checksums(d):
+    lines = []
+    for p in sorted(d.rglob("*")):
+        if p.is_file() and p.name not in ("CHECKSUMS.sha256", "README.md") and "images" not in p.parts:
+            lines.append(f"{sha256(p)}  {p.relative_to(d)}")
+    atomic_write(d / "CHECKSUMS.sha256", lambda t: t.write_text("\n".join(lines) + "\n"))
+
+
+# --------------------------------------------------------------------------
+# Camera traps
+# --------------------------------------------------------------------------
+def scrub_csv_copy(src, dst):
+    """Copy a CSV, normalising provenance tokens of the camera derivative."""
+    text = Path(src).read_text()
+    for a, b in ARRIVAL_TOKEN_REWRITES:
+        text = text.replace(a, b)
+    Path(dst).write_text(text)
+
+
+ARRIVAL_TOKEN_REWRITES = [
+    ("_arrival_icde_camera_", "_arrival_camera_"),
+    ("one retained ICDE SAM3-geometry camera Occurrence", "one retained original-SAM3-geometry camera Occurrence"),
+    ("icde_sam3_geometry_only_derivative_20261108", "camera_release_20261108"),
+]
+
+
+def camera_deployments(tables, images):
+    """Per-deployment table: sensor configuration joined to the geolocation
+    config and the number of processed images on disk."""
+    c1 = tables["Camera Trap Configurations / Campaign 1: June 30 – July 3, 2025"].assign(campaign="thewilds_20250630_20250703")
+    c2 = tables["Camera Trap Configurations / Campaign 2: August 4 – August 28, 2025"].assign(campaign="thewilds_20250804_20250828")
+    dep = pd.concat([t for t in (c1, c2) if "Camera" in t.columns], ignore_index=True)
+    dep.insert(0, "deploymentID", dep.Site.str.extract(r"^(TW\d+)")[0] + "-" + dep.Camera)
+    geo = read(SNAP / "scripts/georeferencing/camera_geolocation_config.csv")
+    geo["camera_id"] = geo.camera_id.str.replace("_", "-")  # raw dir is TW09_CT02
+    dep = dep.merge(geo[["collection", "camera_id", "path_contains", "camera_lat", "camera_lon", "hfov_deg", "bearing_deg",
+                         "image_rotation_deg", "target_height_m", "config_source", "notes"]].rename(
+                        columns={"notes": "geolocation_notes"}),
+                    left_on=["campaign", "deploymentID"], right_on=["collection", "camera_id"], how="left"
+                    ).drop(columns=["collection", "camera_id"]).fillna({"path_contains": ""})
+    dep["images_on_disk"] = dep.apply(lambda r: int(((images.dataset == r.campaign) &
+                                                      (images.camera_id.str.replace("_", "-") == r.deploymentID)
+                                                      & (images.release_path.str.contains(r.path_contains, regex=False)
+                                                         if r.path_contains else True)).sum()), axis=1)
+    # A path-specific override (TW09-CT02 SD23, after the Aug 18 move) takes its
+    # images away from the deployment's default row.
+    for (camp, did), g in dep.groupby(["campaign", "deploymentID"]):
+        if len(g) > 1:
+            default = g.index[g.path_contains == ""]
+            dep.loc[default, "images_on_disk"] -= g.loc[g.path_contains != "", "images_on_disk"].sum()
+    dep["released"] = (dep.deploymentID != "TW07-CT05").map({True: "true", False: "false"})
+    return dep
+
+
+def build_camera(out, tables):
+    """Camera tables come from the validated post-triage derivative: the final
+    Event, Occurrence, image-inventory, media-manifest, taxonomy,
+    annotation-provenance, and arrival-analysis products."""
+    src = CAMERA_SOURCE / "smartwilds-camera-traps"
+    ev = read(src / "data/events.csv")
+    oc = read(src / "data/occurrences.csv")
+    images = read(src / "data/images.csv")
+    if len(ev) != 19785 or len(oc) != 72775 or len(images) != 235778:
+        raise RuntimeError("validated camera source count mismatch")
+    if set(oc.geometryProvenance) != {"SAM3"}:
+        raise RuntimeError("validated camera source contains non-SAM3 geometry")
+    if "behavior" in oc.columns or oc.astype(str).apply(
+            lambda col: col.str.contains(r"water|shore|aquatic", case=False, regex=True).any()).any():
+        raise RuntimeError("validated camera source retains water-association data")
+    if oc.associatedMedia.str.startswith("/").any():
+        raise RuntimeError("validated camera source contains absolute media paths")
+    scrubbed = out / REPOS["camera_traps"]
+    for name in ("camera_image_acquisition_arrivals.csv", "camera_operational_occurrence_arrivals.csv"):
+        atomic_write(scrubbed / "analysis" / name, lambda t, n=name: scrub_csv_copy(src / "analysis" / n, t))
+    atomic_write(scrubbed / "metadata/provenance/ANNOTATION_PROVENANCE.md",
+                 lambda t: scrub_text_copy(src / "metadata/provenance/ANNOTATION_PROVENANCE.md", t,
+                                           extra=[("# ICDE camera-trap annotation provenance", "# Camera-trap annotation provenance")]))
+    files = {
+        "data/events.csv": src / "data/events.csv",
+        "data/occurrences.csv": src / "data/occurrences.csv",
+        "data/images.csv": src / "data/images.csv",
+        "media_manifest.csv": src / "media_manifest.csv",
+        "analysis/camera_arrival_statistics.json": src / "analysis/camera_arrival_statistics.json",
+        "metadata/DATA_DICTIONARY.md": src / "metadata/DATA_DICTIONARY.md",
+        "metadata/taxonomy_mapping.csv": src / "metadata/taxonomy_mapping.csv",
+        "metadata/deployments.csv": camera_deployments(tables, images),
+        "metadata/camera_models.csv": tables["Camera Trap Configurations / Camera Models"],
+        "metadata/camera_geolocation_config.csv": SNAP / "scripts/georeferencing/camera_geolocation_config.csv",
+        "metadata/timestamp_summary_by_camera.csv": TS_MAP / "camera_timestamp_summary_by_camera.csv",
+        "metadata/timestamp_summary_by_date.csv": TS_MAP / "camera_timestamp_summary_by_date.csv",
+        "metadata/provenance/tw07_ct05_sensitive_media_disposition_20261008.json":
+            SNAP / "tw07_ct05_sensitive_media_disposition_20261008.json",
+        "scripts/geolocate_camera_trap_treeoflife_predictions.py":
+            PUBLIC_SCRIPTS / "georeferencing/geolocate_camera_trap_treeoflife_predictions.py",
+        "scripts/generate_darwin_core.py": PUBLIC_SCRIPTS / "darwin_core/generate_darwin_core.py",
+    }
+    return files, {
+        "events": len(ev), "occurrences": len(oc), "images_total": len(images),
+        "all_geometry_provenance_is_SAM3": True,
+        "absolute_media_paths": 0,
+        "water_association_attributes_absent": True,
+        "pass": True,
+    }
+
+
+# --------------------------------------------------------------------------
+# Bioacoustics
+# --------------------------------------------------------------------------
+def build_bioacoustics(out, tables):
+    ev = read(SNAP / "bioacoustics_complete/darwin_core_events_bioacoustics_complete.csv")
+    oc = read(SNAP / "bioacoustics_complete/darwin_core_occurrences_bioacoustics_complete.csv")
+    rec = read(SNAP / "bioacoustics_complete/RAW_WAV_RECONCILIATION_20261008.csv")
+    rec.insert(0, "release_path", rec.raw_wav_path.map(lambda p: rel_media(PROJECT_PREFIX + p)))
+    included = rec.release_disposition.str.startswith("included")
+    path_by_event = rec[included].set_index("darwin_core_eventID").release_path
+    # associatedMedia is a bare WAV filename; resolve it through the Event.
+    oc["associatedMedia"] = oc.eventID.map(path_by_event).fillna(oc.associatedMedia)
+    ev["associatedMedia"] = ev.eventID.map(path_by_event).fillna("")
+
+    disposition = dict(zip(rec.release_path, rec.release_disposition))
+    def excl(rel):
+        d = disposition.get(rel)
+        return None if d is None or d.startswith("included") else d.split(";")[0] + ":" + rec.loc[
+            rec.release_path == rel, "reconciliation_status"].iloc[0]
+    manifest = media_inventory("bioacoustics", excl)
+
+    t = {k: v for k, v in tables.items() if k.startswith("Bioacoustic Recorder Configurations / Campaign")}
+    dep = pd.concat([v.assign(campaign=("thewilds_20250630_20250703" if "June" in k else "thewilds_20250804_20250828"))
+                     for k, v in t.items()], ignore_index=True)
+    dep.insert(0, "deploymentID", dep.Site.str.extract(r"^(TW\d+)")[0] + "-" + dep.Recorder)
+    dep.insert(1, "model", "Wildlife Acoustics Song Meter Micro 2")
+    counts = rec.groupby(["source_campaign_directory", "deploymentID"]).release_disposition.agg(
+        wav_paths="size", wav_released=lambda s: s.str.startswith("included").sum()).reset_index()
+    counts["campaign"] = counts.source_campaign_directory.map(
+        {"campaign1": "thewilds_20250630_20250703", "campaign2": "thewilds_20250804_20250828"})
+    dep = dep.merge(counts.drop(columns="source_campaign_directory"), on=["campaign", "deploymentID"], how="left")
+    birdnet = pd.DataFrame([{"software": "BirdNET Analyzer", "version": "2.4", "segment_length_s": 3.0,
+                             "sample_rate_hz": 48000, "bandpass_hz": "0-15000",
+                             "occurrence_unit": "one row per recording x species; individualCount = detections"}])
+    ev, oc = strip_cluster_paths(ev), strip_cluster_paths(oc)
+    res = check_tables("bioacoustics", ev, oc, manifest, extra_event_ids={"TheWilds_2025"})
+    files = {
+        "data/events.csv": ev, "data/occurrences.csv": oc, "data/recordings.csv": rec,
+        "media_manifest.csv": manifest,
+        "metadata/deployments.csv": dep, "metadata/birdnet_settings.csv": birdnet,
+        "metadata/provenance/RAW_WAV_RECONCILIATION_20261008.md":
+            SNAP / "bioacoustics_complete/RAW_WAV_RECONCILIATION_20261008.md",
+        "metadata/provenance/bioacoustic_event_repair_20261008.json": SNAP / "bioacoustic_event_repair_20261008.json",
+        "scripts/reconcile_bioacoustic_raw_wavs.py": PUBLIC_SCRIPTS / "darwin_core/reconcile_bioacoustic_raw_wavs.py",
+    }
+    return files, res
+
+
+# --------------------------------------------------------------------------
+# Drones
+# --------------------------------------------------------------------------
+def build_drones(out, tables):
+    ev = read(SNAP / "drones/darwin_core_events_drones.csv")
+    oc = read(SNAP / "drones/darwin_core_occurrences_drones.csv")
+    manifest = media_inventory("drones")
+    videos = manifest[manifest.release_path.str.lower().str.endswith((".mp4", ".mov"))]
+    by_stem = {}
+    for p in videos.release_path:
+        flight_dir = p.split("/")[2]
+        by_stem.setdefault((flight_dir, Path(p).stem), []).append(p)
+
+    # Defect 21: the manual starling supplement was appended twice. Drop exact
+    # duplicate rows so every occurrenceID is unique.
+    before = len(oc)
+    oc = oc.drop_duplicates().reset_index(drop=True)
+    rep = {"exact_duplicate_rows_dropped": before - len(oc),
+           "duplicate_occurrenceIDs_remaining": int(oc.occurrenceID.duplicated().sum())}
+    # Frame images (purged scratch frames, or frames exported to an external
+    # annotation workspace) are not shipped; fall back to the retained video.
+    frame_re = re.compile(r"/drones/([^/]+)/.*?/([^/]+)_frames/[^/]+$")
+    def video_for(frame):
+        m = frame_re.search(frame)
+        if not m:
+            return None
+        hits = by_stem.get((m.group(1), m.group(2)), [])
+        return hits[0] if len(hits) == 1 else None
+    frames = oc.associatedMedia.str.startswith("/") & oc.associatedMedia.map(lambda p: rel_media(p) is None)
+    fallback = oc.loc[frames, "associatedMedia"].map(video_for)
+    rep.update({"frame_rows": int(frames.sum()),
+                "frame_rows_unique": int(oc.loc[frames, "associatedMedia"].nunique()),
+                "frame_rows_unmapped": int(fallback.isna().sum())})
+    ok = fallback.dropna().index
+    oc.loc[ok, "associatedMedia"] = fallback[ok]
+    oc.loc[ok, "occurrenceRemarks"] = oc.loc[ok, "occurrenceRemarks"] + \
+        "; associated_media_source=original_video_fallback_frame_not_distributed"
+    oc["associatedMedia"] = oc.associatedMedia.map(lambda p: rel_media(p) or p)
+    # Per-detection table for frame extraction from the released raw videos.
+    # detection_id = <video>_<frame:06d>_<k>; frame is the native frame index
+    # (frames were sampled every 15th frame for SAM3).
+    det = pd.DataFrame({
+        "occurrenceID": oc.occurrenceID, "eventID": oc.eventID, "video_path": oc.associatedMedia,
+        "detection_id": oc.occurrenceRemarks.str.extract(r"detection_id=([^;]+)")[0]})
+    det["frame_index"] = det.detection_id.str.extract(r"_(\d{6})_\d+$")[0].astype("Int64")
+    bbox = oc.identificationRemarks.str.extract(r"bbox xywh\(px\): \[([^\]]+)\]")[0].str.split(",", expand=True)
+    for i, c in enumerate(["bbox_x", "bbox_y", "bbox_w", "bbox_h"]):
+        det[c] = pd.to_numeric(bbox[i]).round(2)
+    det["scientificName"], det["vernacularName"] = oc.scientificName, oc.vernacularName
+    det["top1_score"] = pd.to_numeric(oc.identificationRemarks.str.extract(r"top1 score: ([0-9.eE-]+)")[0])
+    det["georef_status"] = oc.occurrenceRemarks.str.extract(r"drone_georef_status=([^;]+)")[0]
+    det = det[det.detection_id.notna()]  # the 644 curated review records carry no frame id
+    rep["detections_with_frame_index"] = int(det.frame_index.notna().sum())
+    # Remaining frame PNGs that exist on disk are derived products, not raw media.
+    flights = ev[ev.eventType == "SensorDeployment"][["eventID", "eventDate", "locationID", "fieldNotes"]].copy()
+    flights["platform"] = flights.locationID.map(
+        lambda s: "DJI Air 2" if "djiair" in s else "DJI (unspecified airframe)" if "DJI" in s else "Parrot Anafi 4K")
+    media_events = ev[ev.eventType == "MediaEvent"]
+    flights["videos"] = flights.eventID.map(media_events.parentEventID.value_counts()).fillna(0).astype(int)
+    flights["occurrences"] = flights.eventID.map(
+        oc.eventID.map(media_events.set_index("eventID").parentEventID).value_counts()).fillna(0).astype(int)
+    ev, oc = strip_cluster_paths(ev), strip_cluster_paths(oc)
+    res = check_tables("drones", ev, oc, manifest, extra_event_ids={"TheWilds_2025"})
+    res.update(rep)
+    res["pass"] = res["pass"] and rep["duplicate_occurrenceIDs_remaining"] == 0
+    files = {
+        "data/events.csv": ev, "data/occurrences.csv": oc, "data/detections.csv": det,
+        "media_manifest.csv": manifest,
+        "analysis/bird_water_association_per_box.csv": strip_cluster_paths(read(BIRD_WATER)),
+        "metadata/flights.csv": flights, "metadata/platforms.csv": tables["Drone Configurations"],
+        "metadata/drone_projection_config.csv": SNAP / "scripts/georeferencing/drone_projection_config.csv",
+        "metadata/provenance/invalid_drone_date_repair_20261008.json": SNAP / "invalid_drone_date_repair_20261008.json",
+        "metadata/provenance/invalid_drone_coordinate_repair_20261009.json":
+            SNAP / "invalid_drone_coordinate_repair_20261009.json",
+        "metadata/provenance/drone_missing_media_repair_20261010.json": SNAP / "drone_missing_media_repair_20261010.json",
+        "metadata/provenance/excluded_drone_media_event_repair_20261008.json":
+            SNAP / "excluded_drone_media_event_repair_20261008.json",
+        "scripts/geolocate_drone_treeoflife_predictions.py":
+            PUBLIC_SCRIPTS / "georeferencing/geolocate_drone_treeoflife_predictions.py",
+        "scripts/generate_darwin_core.py": PUBLIC_SCRIPTS / "darwin_core/generate_darwin_core.py",
+    }
+    return files, res
+
+
+# --------------------------------------------------------------------------
+# Citizen science and umbrella (need files that are currently owner-only)
+# --------------------------------------------------------------------------
+def build_citizen_science(out, tables):
+    evp = SNAP / "citizen_science/darwin_core_events_citizen_science.csv"
+    ocp = SNAP / "citizen_science/darwin_core_occurrences_citizen_science.csv"
+    if not (readable(evp) and readable(ocp)):
+        raise PermissionError(f"unreadable: {evp.name}, {ocp.name} (owner-only permissions)")
+    ev, oc = read(evp), read(ocp)
+    oc["associatedMedia"] = oc.associatedMedia.map(lambda p: rel_media(p) or p)
+    # The Sep 8 Google Photos album export is not in the project archive, so its
+    # 50 photos cannot be distributed: keep the Occurrences, clear the link, say why.
+    album = oc.associatedMedia.str.startswith("Google Photos Takeout/")
+    oc.loc[album, "occurrenceRemarks"] = oc.loc[album, "occurrenceRemarks"] + \
+        "; media_not_distributed=Google Photos album export not retained in project archive (was " + \
+        oc.loc[album, "associatedMedia"] + ")"
+    oc.loc[album, "associatedMedia"] = ""
+    manifest = media_inventory("citizen_science")
+    ev, oc = strip_cluster_paths(ev), strip_cluster_paths(oc)
+    res = check_tables("citizen_science", ev, oc, manifest, extra_event_ids={"TheWilds_2025"})
+    res["occurrences_media_not_distributed"] = int(album.sum())
+    return {"data/events.csv": ev, "data/occurrences.csv": oc, "media_manifest.csv": manifest,
+            "metadata/provenance/citizen_science_media_reconciliation_20261008.json":
+                SNAP / "citizen_science_media_reconciliation_20261008.json"}, res
+
+
+def build_core(out, tables):
+    """Study-level dataset: validated merged tables, text records, sensor
+    configurations, supplement, defect inventory, and the arrival traces."""
+    src = CAMERA_SOURCE / "smartwilds/data"
+    ev = read(src / "events_all_modalities.csv")
+    oc = read(src / "occurrences_all_modalities.csv")
+    if len(ev) != 22242 or len(oc) != 401602 or oc.occurrenceID.duplicated().any():
+        raise RuntimeError("validated merged-table integrity check failed")
+    # The merged table carries the drone and citizen-science rows with their
+    # archive media paths. Take the release-relative values from the staged
+    # modality tables (same occurrenceID), then rewrite anything left over.
+    rep = {}
+    for repo in ("smartwilds-drones", "smartwilds-citizen-science"):
+        p = out / repo / "data/occurrences.csv"
+        if not p.exists():
+            rep[f"{repo}_not_staged"] = True
+            continue
+        m = read(p).drop_duplicates("occurrenceID").set_index("occurrenceID")
+        idx = oc.occurrenceID.isin(m.index)
+        for c in ("associatedMedia", "occurrenceRemarks", "associatedReferences"):
+            if c in m.columns and c in oc.columns:
+                oc.loc[idx, c] = oc.loc[idx, "occurrenceID"].map(m[c]).values
+        rep[f"{repo}_rows_aligned"] = int(idx.sum())
+    oc["associatedMedia"] = oc.associatedMedia.map(lambda p: rel_media(p) or p)
+    ev, oc = strip_cluster_paths(ev), strip_cluster_paths(oc)
+    res = {"events": len(ev), "occurrences": len(oc),
+           "absolute_media_paths": int(oc.associatedMedia.str.startswith("/").sum()),
+           "cells_with_cluster_paths": count_cluster_paths(ev) + count_cluster_paths(oc)}
+    res.update(rep)
+    res["pass"] = res["absolute_media_paths"] == 0 and res["cells_with_cluster_paths"] == 0
+    files = {
+        "data/events_all_modalities.csv": ev,
+        "data/occurrences_all_modalities.csv": oc,
+        "data/text_events.csv": SNAP / "text_data/darwin_core_events_text.csv",
+        "data/text_occurrences.csv": SNAP / "text_data/darwin_core_occurrences_text.csv",
+        "data/site_selection_rationale.csv": SNAP / "text_data/site_selection_rationale.csv",
+        "metadata/sensor_configurations.md": SENSOR_MD,
+        "SUPPLEMENTAL.md": CARDS / "SUPPLEMENTAL.md",
+        "metadata/DEFECT_INVENTORY.md": CARDS / "DEFECT_INVENTORY.md",
+        "metadata/defect_inventory.csv": CARDS / "defect_inventory.csv",
+        "scripts/refresh_camera_arrival_metadata.py": ROOT / "release/scripts/refresh_camera_arrival_metadata.py",
+        "scripts/generate_arrival_traces_v2.py": PUBLIC_SCRIPTS / "arrival_traces/generate_arrival_traces_v2.py",
+    }
+    # Non-camera arrival traces come from the integration pass; the camera
+    # streams are refreshed afterwards from the validated derivative. The two
+    # notes about the first-generation camera trace no longer apply.
+    atomic_write(out / REPOS["core"] / "metadata/provenance/DARWIN_CORE_AUDIT_20261010_FINAL.json",
+                 lambda t: scrub_text_copy(SNAP / "DARWIN_CORE_AUDIT_20261010_FINAL.json", t))
+    skip = {"DEPRECATED_LEGACY_PRIMITIVES.md", "UNAVAILABLE_CAMERA_IMAGE_CAPTURE_ACQUISITION.md"}
+    drone_media = None
+    for p in sorted((SNAP / "analysis/arrival_traces_v2").iterdir()):
+        if p.name in skip or p.name.startswith("camera_"):
+            continue
+        if p.suffix == ".csv":
+            tr = read(p)
+            if tr.associatedMedia.str.startswith("/").any():
+                if drone_media is None:
+                    drone_media = oc.drop_duplicates("occurrenceID").set_index("occurrenceID").associatedMedia
+                by_source = tr.sourceID.map(drone_media)
+                tr["associatedMedia"] = by_source.where(by_source.notna(), tr.associatedMedia.map(lambda x: rel_media(x) or x))
+            tr = strip_cluster_paths(tr)
+            if "sourceSnapshot" in tr.columns:
+                tr["sourceSnapshot"] = tr.sourceSnapshot.str.replace(SNAP.name, "integration_pass_20261022", regex=False)
+            res[f"arrival_trace_cluster_paths:{p.name}"] = count_cluster_paths(tr)
+            res["pass"] = res["pass"] and res[f"arrival_trace_cluster_paths:{p.name}"] == 0
+            files[f"analysis/arrival_traces_v2/{p.name}"] = tr
+        else:
+            files[f"analysis/arrival_traces_v2/{p.name}"] = p
+    return files, res
+
+
+def main():
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--out", default=str(RELEASE_DIR))
+    ap.add_argument("--only", nargs="*", choices=list(REPOS))
+    args = ap.parse_args()
+    out = Path(args.out)
+    out.mkdir(parents=True, exist_ok=True)
+    tables = parse_md_tables(SENSOR_MD)
+    audit["built_utc"] = datetime.now(timezone.utc).isoformat(timespec="seconds")
+    builders = {"camera_traps": build_camera, "bioacoustics": build_bioacoustics, "drones": build_drones,
+                "citizen_science": build_citizen_science, "core": build_core}
+    for key, fn in builders.items():
+        if args.only and key not in args.only:
+            continue
+        repo = REPOS[key]
+        print(f"[{repo}] building ...", flush=True)
+        for rel in STALE:
+            (out / repo / rel).unlink(missing_ok=True)
+        card = CARDS / f"{repo}.md"
+        if card.exists():
+            atomic_write(out / repo / "README.md", lambda t: shutil.copyfile(card, t))
+        imgs = CARDS / "images" / repo
+        if imgs.is_dir():
+            # Release-card images are static assets. Avoid rewriting existing
+            # owner-managed media assets during an in-place table rebuild.
+            target_images = out / repo / "images"
+            if not target_images.exists():
+                shutil.copytree(imgs, target_images)
+        try:
+            files, res = fn(out, tables)
+        except PermissionError as e:
+            audit["blocked"].append({"repo": repo, "reason": str(e)})
+            print(f"[{repo}] BLOCKED: {e}", flush=True)
+            continue
+        files[f"scripts/{Path(__file__).name}"] = Path(__file__)
+        d = write_repo(out, repo, files)
+        if key != "camera_traps":
+            write_dictionary(d, repo)
+        write_checksums(d)
+        audit["repos"][repo] = res
+        print(f"[{repo}] {json.dumps({k: v for k, v in res.items() if not isinstance(v, (list, dict))})}",
+              flush=True)
+    # The camera arrival streams in the study-level dataset come from the
+    # validated camera derivative; refresh them once both parts are staged.
+    requested = set(args.only or REPOS)
+    if {"camera_traps", "core"}.issubset(requested):
+        subprocess.run([sys.executable, str(ROOT / "release/scripts/refresh_camera_arrival_metadata.py"),
+                        "--release", str(out), "--snapshot", str(SNAP), "--source", str(CAMERA_SOURCE)], check=True)
+        write_dictionary(out / REPOS["core"], REPOS["core"])
+        write_checksums(out / REPOS["core"])
+    (out / "build_hf_release.py").unlink(missing_ok=True)  # stray copy from an earlier build
+    (out / "RELEASE_BUILD_AUDIT.json").write_text(json.dumps(audit, indent=2, default=str) + "\n")
+    return 0 if not audit["errors"] else 1
+
+
+if __name__ == "__main__":
+    sys.exit(main())

smartwilds-drones/CHECKSUMS.sha256

@@ -0,0 +1,17 @@
+690d9c85b2c2ff71df456ef4d27edd0b741aa168fa9e5cdb74cfb2a31cf389dd  analysis/bird_water_association_per_box.csv
+f54c1f55c771dab92cb04eca38456c53263df4f88389a0cf1d141ed89a95bfdb  data/detections.csv
+b080e3fa5a46d150717a9e28a355eb8654c6878d5451c6f7b0f19fa15a574085  data/events.csv
+0ffefa18d45de1457f4c27ad6da6a21f483cbf8c3e64e59cda3d620e1571c6bb  data/occurrences.csv
+08ef9e0b5c73f1ddeb1355b20405970f9d3e083063c6207b2adcdc0680e9ea3b  media_manifest.csv
+2ed12997a39b04bfb5615cb4c5cc46a1256bd3f862e80d1807c97c45fbbf57f3  metadata/DATA_DICTIONARY.md
+700405c4097f216c8a92a9052d82a6f1eef4a77ad70f5cb32628eb8ab9804368  metadata/drone_projection_config.csv
+184070fe638c4f794529ec11be9f8480f1b5b0e667ee4f59d7f6c727407cf2ae  metadata/flights.csv
+db48200f39c4bb82d5f88449705e46086f3c8ef2034e43e1483a0912d1c32518  metadata/platforms.csv
+38900fec3c4963585e3fef855e2e22097594a3fd575815aa34fb9ee023fc962e  metadata/provenance/drone_missing_media_repair_20261010.json
+28094bef545cef01ed1d281746a77ce7f3559527188931ac5788ad1cb50a6d4b  metadata/provenance/drone_time_recovery_20250811.json
+12b3b4efecc1fdcffd2f6a077277fb5b534953298ea54bf7a0287a0f2c10ed0a  metadata/provenance/excluded_drone_media_event_repair_20261008.json
+6d8eaa386f91ecbf45583e3985ba14c5c2eddf1132201bb72a2afeaf134b6124  metadata/provenance/invalid_drone_coordinate_repair_20261009.json
+1e2b46676692871b7c30cf3fca6607c823709f48964b60246a84fea12d796de0  metadata/provenance/invalid_drone_date_repair_20261008.json
+f2efc6692a5d7add094933b71b4f64a15efa2fd1c2578bb66620ff161747934a  scripts/build_hf_release.py
+6e90073a9d7bcaf1c6d37fbede20d7444d07792000888c8d729e5787986d9545  scripts/generate_darwin_core.py
+9f08e2e4d73c737357b5f3b513baa52be31cee49c394058cf05448b1401f6513  scripts/geolocate_drone_treeoflife_predictions.py

smartwilds-drones/README.md

@@ -0,0 +1,305 @@
+---
+license: cc0-1.0
+language:
+- en
+pretty_name: SmartWilds Drones (The Wilds 2025)
+task_categories:
+- object-detection
+- image-classification
+tags:
+- biology
+- ecology
+- biodiversity
+- drone
+- uav
+- aerial-imagery
+- darwin-core
+- humboldt-eco
+- fair2
+- smartwilds
+- the-wilds
+- ohio
+size_categories:
+- 100K<n<1M
+description: >-
+  39 drone videos (95 GB) from 15 flight sessions on 6 dates at The Wilds
+  (Ohio, USA), July-September 2025, from three platforms (Parrot Anafi 4K, DJI,
+  DJI Air 2), with flight telemetry and 318,215 Darwin Core Occurrences
+  (frame-level SAM3 + BioCLIP Tree-of-Life detection crops with projected
+  ground positions).
+
+fair2_compliance:
+  findable:
+    doi: ""  # pending the raw data release
+    metadata_registry: ["DataCite", "GBIF"]
+  accessible:
+    open_access: true
+    authentication_required: false
+  interoperable:
+    standards: ["Darwin Core", "Humboldt Eco", "FAIR2"]
+  reusable:
+    license_clear: true
+    provenance_documented: true
+  ai_ready:
+    machine_readable: true
+    structured_annotations: true
+
+darwin_core:
+  event_coverage:
+    start_date: "2025-07-01"
+    end_date: "2025-09-08"
+    decimal_latitude: 39.83
+    decimal_longitude: -81.74
+    coordinate_uncertainty_meters: 5000
+    locality: "The Wilds, Cumberland, Muskingum County, Ohio, USA"
+  occurrence_info:
+    basis_of_record: "MachineObservation"
+    sampling_protocol: "Manual drone video survey of pastures and lake margin; SAM3 detection on sampled frames, BioCLIP Tree-of-Life classification, ground projection from telemetry"
+
+platform:
+  type: "UAV"
+  models: ["Parrot Anafi 4K", "DJI (airframe unrecorded)", "DJI Air 2"]
+  autonomy_mode: "manual"
+
+sensors:
+  - type: "RGB"
+    platform: "Parrot Anafi 4K"
+    video: "4K MP4"
+    telemetry: "Parrot JSON / GUTMA DX"
+  - type: "RGB"
+    platform: "DJI (airframe unrecorded)"
+    video: "1080p MOV"
+    telemetry: "CSV, 10 Hz"
+  - type: "RGB"
+    platform: "DJI Air 2"
+    video: "4K MP4"
+    telemetry: "Airdata CSV"
+
+mission:
+  telemetry_available: true
+
+configs:
+- config_name: occurrences
+  data_files: data/occurrences.csv
+- config_name: events
+  data_files: data/events.csv
+- config_name: flights
+  data_files: metadata/flights.csv
+---
+
+# Dataset Card for SmartWilds Drones
+
+![Drone detections of three managed focal species](images/smartwilds-drones/preview.png)
+*SAM3 detections on drone frames of Przewalski's horse, Sichuan takin, and Milu (Père David's) deer.*
+
+**Every drone video flown at The Wilds in 2025, with each platform's raw
+telemetry, and frame-level animal detections placed on the ground where the
+telemetry allows it.**
+
+This is the drone dataset of the **SmartWilds** collection. Each modality has
+its own dataset and DOI. They share one Darwin Core Event hierarchy rooted at
+`TheWilds_2025` and are joined by the
+[SmartWilds collection](https://huggingface.co/collections/ABC-Center/smartwilds-6af3c0aac0a67864cbf6edb4).
+
+## Dataset Details
+
+- **Curated by:** Jenna M. Kline, Anirudh Potlapally, Elizabeth G. Campolongo
+- **Homepage:** <https://github.com/Imageomics/smartwilds> (documentation: <https://imageomics.github.io/smartwilds/>)
+- **Paper:** SmartWilds: Integrating, Aligning, and Documenting Multi-Modal Field Sensing Data (under review)
+- **Release build:** 2026-10-22
+
+| Campaign | Dates | Sessions | Platform |
+|---|---|---:|---|
+| 1 | Jul 1, Jul 3 | 6 | Parrot Anafi 4K |
+| 2 | Aug 11, 18, 25 | 4 + 1 | Parrot Anafi 4K; DJI (Aug 11) |
+| 3 | Sep 8 | 4 | DJI Air 2 |
+
+**Darwin Core tables.** The Event table has 55 rows: 1 `Project`, 15
+`SensorDeployment` flight sessions, and 39 `MediaEvent` videos. There are
+318,215 Occurrences: 318,203 machine detection crops on sampled frames and
+12 human-added European starling boxes (`human-supported; limited-resolution`).
+`metadata/flights.csv` summarizes each session with its platform, number of
+videos, and number of Occurrences.
+
+The 15 sessions are the 15 directories that contain video. Three more Aug 25
+Parrot directories contain telemetry but no video. Their files are included in
+`media/`, but they are not sampling Events.
+
+### Supported Tasks
+
+- **Aerial detection and classification:** you can re-label the raw video and
+  compare against the released crops.
+- **Telemetry-aware vision:** altitude, heading, and gimbal pitch are available
+  per frame where the platform logged them.
+- **Workload studies:** detection arrivals are extremely bursty (B ≈ +0.99), so
+  they are a stress case for scheduling and offload.
+
+## Dataset Structure
+
+```
+smartwilds-drones/
+├── data/
+│   ├── events.csv          # Project, 15 flight SensorDeployments, 39 MediaEvents
+│   ├── occurrences.csv     # 318,215 frame-level detection crops
+│   └── detections.csv      # per detection: video path, frame index, box (px)
+├── media/                  # media/<campaign>/<flight>/... videos + telemetry
+├── media_manifest.csv
+├── metadata/
+│   ├── flights.csv         # per-session platform, video and occurrence counts
+│   ├── platforms.csv       # resolution, telemetry format per platform
+│   ├── drone_projection_config.csv
+│   └── provenance/         # date, coordinate, media and excluded-video audits
+├── analysis/bird_water_association_per_box.csv
+└── scripts/
+```
+
+### Raw video, not frames
+
+The release ships the **raw videos**. It does not ship extracted frames. Each
+machine detection in `data/detections.csv` gives the video (`video_path`), the
+native frame number (`frame_index`), and the box in pixels. SAM3 ran on every
+15th frame. Any frame can be regenerated:
+
+```python
+import cv2, pandas as pd
+det = pd.read_csv("data/detections.csv")
+row = det.iloc[0]
+cap = cv2.VideoCapture(row.video_path)
+cap.set(cv2.CAP_PROP_POS_FRAMES, int(row.frame_index))
+ok, frame = cap.read()
+x, y, w, h = map(int, (row.bbox_x, row.bbox_y, row.bbox_w, row.bbox_h))
+crop = frame[max(y, 0):y + h, max(x, 0):x + w]
+```
+
+The 644 curated goose records come from a separate 20-frame review subset and
+have no `frame_index`.
+
+### Annotation method
+
+- **Method:** automated, with a human-validated subset.
+- **Detection:** SAM3 on frames extracted from each video.
+- **Classification:** top-1 label from BioCLIP Tree-of-Life, written to the
+  Darwin Core taxon fields. The rank, score, box, and altitude are in
+  `identificationRemarks`.
+- **Ground projection:** the ray through the bottom-center of the box is
+  intersected with a flat ground plane, using drone GPS, AGL altitude,
+  heading, gimbal pitch (with a fallback), and approximate FOV.
+  `drone_georef_status` in `occurrenceRemarks` gives the result for each
+  detection:
+
+  | Status | Rows |
+  |---|---:|
+  | `ok` | 247,340 |
+  | `invalid_altitude` | 27,306 |
+  | `missing_drone_gps` | 24,467 |
+  | `horizon_ray` | 18,324 |
+  | `out_of_range` | 122 |
+
+  When the status is not `ok`, the coordinate is the drone's own position, a
+  validated static video-level location (500 m uncertainty), or blank.
+  `dataGeneralizations` says which on every row. The flights were flown for
+  population counts and behavior, not geolocation, so platform position
+  places an animal in the right pasture.
+- **Verification:** 317,559 rows are `unverified`. 644 are `validated` Canada
+  goose records from a 20-frame review of `20250703_parrotanafi_5`.
+
+## Dataset Creation
+
+### Known defects and repairs
+
+- **Aug 11 session had no time of day.** Its `eventDate` was exported as the
+  string `nan` (41,524 Occurrences); the date was restored from the flight
+  identifier. The times were recovered from the video container creation
+  times, with each aircraft's clock convention settled against reference
+  flights and the day's field log (`metadata/provenance/drone_time_recovery_20250811.json`).
+  The Parrot times are verified to the second; the DJI times rest on an
+  inferred local-time convention and should be read as minute-level.
+- **Impossible coordinates on 7 Events** (`nan`, or longitude ≈ 500). These were
+  replaced with a validated static video location, or left blank.
+- **Frame media lost from scratch storage.** Extracted frames were deleted when
+  the scratch workspace was purged. The affected Occurrences now point at the
+  original video and are tagged
+  `associated_media_source=original_video_fallback_missing_scratch_frame`.
+  The release build on 2026-10-22 applied the same fallback to the last
+  10,943 rows (479 frames). As a result, every drone `associatedMedia` value
+  resolves to a released video.
+- **One video was not processed.** `20250703_parrotanafi_4/VID_001` was reviewed
+  and found to have no usable content. It is a `MediaEvent` with zero
+  Occurrences.
+- **Three telemetry schemas.** Field names, units (feet vs. meters), and gimbal
+  conventions differ by platform and were mapped per platform.
+
+### Personal and sensitive information
+
+3,114 Occurrences are labeled *Homo sapiens*: field staff and visitors seen
+from altitude. Media in which a person can be individually identified are removed from the
+release. People too small to identify (a person box under 80 px in the native
+image, where the face is under about 10 px) are retained; every file with a
+larger person detection was reviewed by a person before release.
+For drone videos this means trimming launch and landing segments where the
+operators are identifiable; the trims are lossless and recorded in the
+media manifest.
+
+## Considerations for Using the Data
+
+- **Detections count crops, not animals.** Adjacent frames repeat the same
+  animal, so the counts measure evidence volume, not abundance.
+- **Species-rank labels need curation.** The class-rank accuracy reported for
+  the camera-trap subset (0.845) is the better guide to these labels.
+  <!-- TODO(confirm-stats): the camera-trap card no longer reports this BioCLIP class accuracy (its labels are now triage-based). Keep the figure here only if the BioCLIP evaluation is reported elsewhere, e.g. SUPPLEMENTAL.md. -->
+
+- **The DJI airframe for Aug 11 was not recorded.**
+
+## Supplemental Material
+
+Extended inventories, the timestamp-repair methods, the standards crosswalk,
+validation checks, and the full defect inventory are in
+[`SUPPLEMENTAL.md`](https://huggingface.co/datasets/ABC-Center/smartwilds_metadata/blob/main/SUPPLEMENTAL.md)
+in the study-level SmartWilds dataset.
+
+## Licensing
+
+Data: [CC0 1.0](https://creativecommons.org/publicdomain/zero/1.0/). Scripts: MIT.
+
+## Citation
+
+```bibtex
+@misc{smartwilds2025_drones,
+  author    = {Kline, Jenna M. and Potlapally, Anirudh and Campolongo, Elizabeth G.},
+  title     = {SmartWilds Drones (The Wilds 2025)},
+  year      = {2026},
+  publisher = {Hugging Face},
+  note      = {DOI pending the raw data release},
+  url       = {https://huggingface.co/datasets/ABC-Center/smartwilds_metadata/tree/main/smartwilds-drones}
+}
+```
+
+Please also cite the paper:
+
+```bibtex
+@article{kline2026smartwilds,
+  author  = {Kline, Jenna M. and Potlapally, Anirudh and Campolongo, Elizabeth G.},
+  title   = {SmartWilds: Integrating, Aligning, and Documenting Multi-Modal Field Sensing Data},
+  year    = {2026},
+  note    = {Under review}
+}
+```
+
+This dataset card follows the [FAIR² Drones Data Standard](https://imageomics.github.io/fair_drones/)
+(Kline and Campolongo, 2026).
+
+## Acknowledgements
+
+This project is supported by the [Imageomics Institute](https://imageomics.org), which is funded by the US National Science Foundation's Harnessing the Data Revolution (HDR) program under [Award #2118240](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2118240) (Imageomics: A New Frontier of Biological Information Powered by Knowledge-Guided Machine Learning), and by the [AI Institute for Intelligent Cyberinfrastructure with Computational Learning in the Environment (ICICLE)](https://icicle.osu.edu/), which is funded by the US National Science Foundation under [Award #2112606](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2112606).
+
+This work is also supported by the AI and Biodiversity Change (ABC) Global Center. The ABC Global Center is funded by the US National Science Foundation under [Award No. 2330423](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2330423&HistoricalAwards=false) and the Natural Sciences and Engineering Research Council of Canada under [Award No. 585136](https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=782440). This dataset draws on research supported by the Social Sciences and Humanities Research Council.
+
+Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation, Natural Sciences and Engineering Research Council of Canada, or Social Sciences and Humanities Research Council.
+
+Special thanks to Dan Beetem and [The Wilds](https://www.thewilds.org/) in Ohio for providing access to deployment sites and supporting this wildlife monitoring research. We thank the Columbus Zoo and Aquarium for their support in facilitating this project. All data collection, including drone flights, was conducted under the supervision of the Director of Animal Management, with permission from The Wilds Animal Care and Use Committee.
+
+This work used the [Ohio Supercomputer Center](https://www.osc.edu/) (Ohio Supercomputer Center. 1987. Ohio Supercomputer Center. Columbus OH: Ohio Supercomputer Center. <https://ror.org/01apna436>).
+
+## Dataset Card Contact
+
+Jenna M. Kline (kline.377@osu.edu)

smartwilds-drones/data/events.csv

@@ -0,0 +1,56 @@
+eventID,parentEventID,eventType,eventDate,samplingProtocol,sampleSizeValue,sampleSizeUnit,locality,locationID,decimalLatitude,decimalLongitude,coordinateUncertaintyInMeters,geodeticDatum,country,stateProvince,county,datasetID,datasetName,recordedBy,fieldNotes,eventTime
+TheWilds_2025_drones,TheWilds_2025,Project,2025-07-01/2025-09-08,"Drone video surveys; detections by SAM3, Tree-of-Life classification by BioCLIP, and first-pass animal ground-position projection from drone telemetry.",39,raw drone videos (38 processed; 1 intentionally excluded),"The Wilds Conservation Center, Muskingum County, Ohio, USA",,39.83,-81.74,5000,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,38 videos regenerated from current Tree-of-Life outputs; VID_001 from 20250703_parrotanafi_4 is retained as an intentionally excluded raw-media Event.,
+TheWilds_2025_drone_20250701_parrotanafi_1,TheWilds_2025_drones,SensorDeployment,2025-07-01,Drone flight/session 20250701_parrotanafi_1; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,,detections,The Wilds Conservation Center,20250701_parrotanafi_1,39.82887083333163,-81.74238066666636,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250701_parrotanafi_1,
+TheWilds_2025_drone_20250703_parrotanafi_1,TheWilds_2025_drones,SensorDeployment,2025-07-03,Drone flight/session 20250703_parrotanafi_1; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,,detections,The Wilds Conservation Center,20250703_parrotanafi_1,39.83069739991588,-81.73770322014592,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_1,
+TheWilds_2025_drone_20250703_parrotanafi_2,TheWilds_2025_drones,SensorDeployment,2025-07-03,Drone flight/session 20250703_parrotanafi_2; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,,detections,The Wilds Conservation Center,20250703_parrotanafi_2,39.83535716666588,-81.73001549999995,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_2,
+TheWilds_2025_drone_20250703_parrotanafi_3,TheWilds_2025_drones,SensorDeployment,2025-07-03,Drone flight/session 20250703_parrotanafi_3; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,,detections,The Wilds Conservation Center,20250703_parrotanafi_3,39.83437116666848,-81.73930916666667,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_3,
+TheWilds_2025_drone_20250703_parrotanafi_4,TheWilds_2025_drones,SensorDeployment,2025-07-03,Drone flight/session 20250703_parrotanafi_4; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,,detections,The Wilds Conservation Center,20250703_parrotanafi_4,39.830678500000175,-81.73834333333349,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_4,
+TheWilds_2025_drone_20250703_parrotanafi_5,TheWilds_2025_drones,SensorDeployment,2025-07-03,Drone flight/session 20250703_parrotanafi_5; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,,detections,The Wilds Conservation Center,20250703_parrotanafi_5,,,,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_5; coordinate_source=unavailable_no_single_session_location,
+TheWilds_2025_drone_20250811_DJI_01,TheWilds_2025_drones,SensorDeployment,2025-08-11,Drone flight/session 20250811_DJI_01; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,,detections,The Wilds Conservation Center,20250811_DJI_01,39.834171,-81.73782,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250811_DJI_01; eventDate derived from flight identifier 20250811; eventTime recovered from the video container creation time (release build 2026-11-10; metadata/provenance/drone_time_recovery_20250811.json),10:29:39/10:35:44
+TheWilds_2025_drone_20250811_parrot_01,TheWilds_2025_drones,SensorDeployment,2025-08-11,Drone flight/session 20250811_parrot_01; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,,detections,The Wilds Conservation Center,20250811_parrot_01,,,,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250811_parrot_01; eventDate derived from flight identifier 20250811; eventTime recovered from the video container creation time (release build 2026-11-10; metadata/provenance/drone_time_recovery_20250811.json); coordinate_source=unavailable_no_single_session_location,09:12:43/09:23:43
+TheWilds_2025_drone_20250818_parrot_01,TheWilds_2025_drones,SensorDeployment,2025-08-18,Drone flight/session 20250818_parrot_01; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,,detections,The Wilds Conservation Center,20250818_parrot_01,39.83062300000001,-81.7380009999996,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250818_parrot_01,
+TheWilds_2025_drone_20250818_parrot_02,TheWilds_2025_drones,SensorDeployment,2025-08-18,Drone flight/session 20250818_parrot_02; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,,detections,The Wilds Conservation Center,20250818_parrot_02,39.83553983333299,-81.729999,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250818_parrot_02,
+TheWilds_2025_drone_20250825_parrot_01,TheWilds_2025_drones,SensorDeployment,2025-08-25,Drone flight/session 20250825_parrot_01; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,,detections,The Wilds Conservation Center,20250825_parrot_01,39.8306344999969,-81.73801366666375,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250825_parrot_01,
+TheWilds_2025_drone_20250908_djiair_1,TheWilds_2025_drones,SensorDeployment,2025-09-08,Drone flight/session 20250908_djiair_1; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,,detections,The Wilds Conservation Center,20250908_djiair_1,39.8306117686427,-81.7381286438897,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250908_20250908; flight/session: 20250908_djiair_1,
+TheWilds_2025_drone_20250908_djiair_2,TheWilds_2025_drones,SensorDeployment,2025-09-08,Drone flight/session 20250908_djiair_2; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,,detections,The Wilds Conservation Center,20250908_djiair_2,39.8306056257256,-81.7381547387215,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250908_20250908; flight/session: 20250908_djiair_2,
+TheWilds_2025_drone_20250908_djiair_3,TheWilds_2025_drones,SensorDeployment,2025-09-08,Drone flight/session 20250908_djiair_3; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,,detections,The Wilds Conservation Center,20250908_djiair_3,39.8341484494756,-81.7376652256038,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250908_20250908; flight/session: 20250908_djiair_3,
+TheWilds_2025_drone_20250908_djiair_4,TheWilds_2025_drones,SensorDeployment,2025-09-08,Drone flight/session 20250908_djiair_4; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,,detections,The Wilds Conservation Center,20250908_djiair_4,39.834147453031,-81.7377155569067,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250908_20250908; flight/session: 20250908_djiair_4,
+TheWilds_2025_drone_20250701_parrotanafi_1_VID_001,TheWilds_2025_drone_20250701_parrotanafi_1,MediaEvent,2025-07-01,Drone video VID_001; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,12,classified detection crops,The Wilds Conservation Center,20250701_parrotanafi_1,39.82887083333163,-81.74238066666636,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250701_parrotanafi_1; video: VID_001; detections: 12; source CSV: thewilds_20250630_20250703_20250701_parrotanafi_1_VID_001.drone_geolocated.csv,08:34:41/08:34:46
+TheWilds_2025_drone_20250701_parrotanafi_1_VID_002,TheWilds_2025_drone_20250701_parrotanafi_1,MediaEvent,2025-07-01,Drone video VID_002; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,5664,classified detection crops,The Wilds Conservation Center,20250701_parrotanafi_1,39.82887183333197,-81.7423786666653,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250701_parrotanafi_1; video: VID_002; detections: 5664; source CSV: thewilds_20250630_20250703_20250701_parrotanafi_1_VID_002.drone_geolocated.csv,08:34:46/08:40:27
+TheWilds_2025_drone_20250701_parrotanafi_1_VID_003,TheWilds_2025_drone_20250701_parrotanafi_1,MediaEvent,2025-07-01,Drone video VID_003; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,6923,classified detection crops,The Wilds Conservation Center,20250701_parrotanafi_1,39.8287242703832,-81.74119786203082,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250701_parrotanafi_1; video: VID_003; detections: 6923; source CSV: thewilds_20250630_20250703_20250701_parrotanafi_1_VID_003.drone_geolocated.csv,08:40:27/08:46:04
+TheWilds_2025_drone_20250701_parrotanafi_1_VID_004,TheWilds_2025_drone_20250701_parrotanafi_1,MediaEvent,2025-07-01,Drone video VID_004; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,3929,classified detection crops,The Wilds Conservation Center,20250701_parrotanafi_1,39.828693882055184,-81.74121949089592,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250701_parrotanafi_1; video: VID_004; detections: 3929; source CSV: thewilds_20250630_20250703_20250701_parrotanafi_1_VID_004.drone_geolocated.csv,08:46:05/08:48:25
+TheWilds_2025_drone_20250703_parrotanafi_1_VID_001,TheWilds_2025_drone_20250703_parrotanafi_1,MediaEvent,2025-07-03,Drone video VID_001; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,380,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_1,39.83069739991588,-81.73770322014592,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_1; video: VID_001; detections: 380; source CSV: thewilds_20250630_20250703_20250703_parrotanafi_1_VID_001.drone_geolocated.csv,15:48:01/15:50:47
+TheWilds_2025_drone_20250703_parrotanafi_1_VID_002,TheWilds_2025_drone_20250703_parrotanafi_1,MediaEvent,2025-07-03,Drone video VID_002; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,41130,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_1,39.83063966666832,-81.73807449999214,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_1; video: VID_002; detections: 41130; source CSV: thewilds_20250630_20250703_20250703_parrotanafi_1_VID_002.drone_geolocated.csv,15:50:59/15:56:33
+TheWilds_2025_drone_20250703_parrotanafi_1_VID_003,TheWilds_2025_drone_20250703_parrotanafi_1,MediaEvent,2025-07-03,Drone video VID_003; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,588,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_1,39.82954685533496,-81.7406838660845,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_1; video: VID_003; detections: 588; source CSV: thewilds_20250630_20250703_20250703_parrotanafi_1_VID_003.drone_geolocated.csv,15:56:34/15:58:18
+TheWilds_2025_drone_20250703_parrotanafi_2_VID_001,TheWilds_2025_drone_20250703_parrotanafi_2,MediaEvent,2025-07-03,Drone video VID_001; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,34,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_2,39.83535716666588,-81.73001549999995,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_2; video: VID_001; detections: 34; source CSV: thewilds_20250630_20250703_20250703_parrotanafi_2_VID_001.drone_geolocated.csv,16:21:04/16:21:10
+TheWilds_2025_drone_20250703_parrotanafi_2_VID_002,TheWilds_2025_drone_20250703_parrotanafi_2,MediaEvent,2025-07-03,Drone video VID_002; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,5933,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_2,39.83536710341573,-81.73004209188048,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_2; video: VID_002; detections: 5933; source CSV: thewilds_20250630_20250703_20250703_parrotanafi_2_VID_002.drone_geolocated.csv,16:21:11/16:26:45
+TheWilds_2025_drone_20250703_parrotanafi_2_VID_003,TheWilds_2025_drone_20250703_parrotanafi_2,MediaEvent,2025-07-03,Drone video VID_003; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,3843,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_2,39.83656393622655,-81.73010565554823,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_2; video: VID_003; detections: 3843; source CSV: thewilds_20250630_20250703_20250703_parrotanafi_2_VID_003.drone_geolocated.csv,16:26:46/16:31:56
+TheWilds_2025_drone_20250703_parrotanafi_3_VID_001,TheWilds_2025_drone_20250703_parrotanafi_3,MediaEvent,2025-07-03,Drone video VID_001; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,169,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_3,39.83437116666848,-81.73930916666667,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_3; video: VID_001; detections: 169; source CSV: thewilds_20250630_20250703_20250703_parrotanafi_3_VID_001.drone_geolocated.csv,16:56:20/16:56:41
+TheWilds_2025_drone_20250703_parrotanafi_3_VID_002,TheWilds_2025_drone_20250703_parrotanafi_3,MediaEvent,2025-07-03,Drone video VID_002; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,523,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_3,39.83434777806453,-81.73937466319438,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_3; video: VID_002; detections: 523; source CSV: thewilds_20250630_20250703_20250703_parrotanafi_3_VID_002.drone_geolocated.csv,16:56:41/17:01:29
+TheWilds_2025_drone_20250703_parrotanafi_4_VID_002,TheWilds_2025_drone_20250703_parrotanafi_4,MediaEvent,2025-07-03,Drone video VID_002; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,7368,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_4,39.830678500000175,-81.73834333333349,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_4; video: VID_002; detections: 7368; source CSV: thewilds_20250630_20250703_20250703_parrotanafi_4_VID_002.drone_geolocated.csv,17:23:46/17:29:16
+TheWilds_2025_drone_20250703_parrotanafi_4_VID_003,TheWilds_2025_drone_20250703_parrotanafi_4,MediaEvent,2025-07-03,Drone video VID_003; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,511,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_4,39.83081839120959,-81.7383211402393,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_4; video: VID_003; detections: 511; source CSV: thewilds_20250630_20250703_20250703_parrotanafi_4_VID_003.drone_geolocated.csv,17:29:25/17:32:36
+TheWilds_2025_drone_20250703_parrotanafi_5_VID_001,TheWilds_2025_drone_20250703_parrotanafi_5,MediaEvent,2025-07-03,Drone video VID_001; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,20,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_5,,,,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_5; video: VID_001; detections: 20; source CSV: thewilds_20250630_20250703_20250703_parrotanafi_5_VID_001.drone_geolocated.csv; coordinate_source=unavailable_no_single_session_location,17:25:55/17:25:59
+TheWilds_2025_drone_20250703_parrotanafi_5_VID_002,TheWilds_2025_drone_20250703_parrotanafi_5,MediaEvent,2025-07-03,Drone video VID_002; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,8383,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_5,39.83024333,-81.73892500,500,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_5; video: VID_002; detections: 8383; source CSV: thewilds_20250630_20250703_20250703_parrotanafi_5_VID_002.drone_geolocated.csv; coordinate_source=original_video_quicktime_location_static; not per-frame telemetry,17:25:59/17:31:34
+TheWilds_2025_drone_20250703_parrotanafi_5_VID_003,TheWilds_2025_drone_20250703_parrotanafi_5,MediaEvent,2025-07-03,Drone video VID_003; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,21437,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_5,39.82953921836771,-81.74022920437498,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_5; video: VID_003; detections: 21437; source CSV: thewilds_20250630_20250703_20250703_parrotanafi_5_VID_003.drone_geolocated.csv,17:31:34/17:35:07
+TheWilds_2025_drone_20250703_parrotanafi_5_VID_004,TheWilds_2025_drone_20250703_parrotanafi_5,MediaEvent,2025-07-03,Drone video VID_004; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,33677,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_5,39.82928278702048,-81.74047635534833,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_5; video: VID_004; detections: 33033; source CSV: thewilds_20250630_20250703_20250703_parrotanafi_5_VID_004.drone_geolocated.csv Includes 644 curated bird occurrence records.,17:35:07/17:40:42
+TheWilds_2025_drone_20250703_parrotanafi_5_VID_005,TheWilds_2025_drone_20250703_parrotanafi_5,MediaEvent,2025-07-03,Drone video VID_005; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,201,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_5,39.82885447083709,-81.74156150249854,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_5; video: VID_005; detections: 201; source CSV: thewilds_20250630_20250703_20250703_parrotanafi_5_VID_005.drone_geolocated.csv,17:40:43/17:42:43
+TheWilds_2025_drone_20250811_DJI_01_DJI_0001,TheWilds_2025_drone_20250811_DJI_01,MediaEvent,2025-08-11,Drone video DJI_0001; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,4535,classified detection crops,The Wilds Conservation Center,20250811_DJI_01,39.834171,-81.73782,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250811_DJI_01; video: DJI_0001; detections: 4535; source CSV: thewilds_20250804_20250828_20250811_DJI_01_DJI_0001.drone_geolocated.csv; eventDate derived from flight identifier 20250811; eventTime recovered from the video container creation time (release build 2026-11-10; metadata/provenance/drone_time_recovery_20250811.json),10:29:39/10:31:17
+TheWilds_2025_drone_20250811_DJI_01_DJI_0002,TheWilds_2025_drone_20250811_DJI_01,MediaEvent,2025-08-11,Drone video DJI_0002; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,12522,classified detection crops,The Wilds Conservation Center,20250811_DJI_01,39.834231,-81.736127,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250811_DJI_01; video: DJI_0002; detections: 12522; source CSV: thewilds_20250804_20250828_20250811_DJI_01_DJI_0002.drone_geolocated.csv; eventDate derived from flight identifier 20250811; eventTime recovered from the video container creation time (release build 2026-11-10; metadata/provenance/drone_time_recovery_20250811.json),10:31:32/10:35:44
+TheWilds_2025_drone_20250811_parrot_01_P5180524,TheWilds_2025_drone_20250811_parrot_01,MediaEvent,2025-08-11,Drone video P5180524; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,41,classified detection crops,The Wilds Conservation Center,20250811_parrot_01,39.83068541,-81.73774815,500,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250811_parrot_01; video: P5180524; detections: 41; source CSV: thewilds_20250804_20250828_20250811_parrot_01_P5180524.drone_geolocated.csv; eventDate derived from flight identifier 20250811; eventTime recovered from the video container creation time (release build 2026-11-10; metadata/provenance/drone_time_recovery_20250811.json); coordinate_source=original_video_quicktime_location_static; not per-frame telemetry,09:12:43/09:14:14
+TheWilds_2025_drone_20250811_parrot_01_P5190525,TheWilds_2025_drone_20250811_parrot_01,MediaEvent,2025-08-11,Drone video P5190525; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,12430,classified detection crops,The Wilds Conservation Center,20250811_parrot_01,39.83034692,-81.73837571,500,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250811_parrot_01; video: P5190525; detections: 12430; source CSV: thewilds_20250804_20250828_20250811_parrot_01_P5190525.drone_geolocated.csv; eventDate derived from flight identifier 20250811; eventTime recovered from the video container creation time (release build 2026-11-10; metadata/provenance/drone_time_recovery_20250811.json); coordinate_source=original_video_quicktime_location_static; not per-frame telemetry,09:14:19/09:19:54
+TheWilds_2025_drone_20250811_parrot_01_P5190526,TheWilds_2025_drone_20250811_parrot_01,MediaEvent,2025-08-11,Drone video P5190526; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,11996,classified detection crops,The Wilds Conservation Center,20250811_parrot_01,39.82935200,-81.74035901,500,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250811_parrot_01; video: P5190526; detections: 11996; source CSV: thewilds_20250804_20250828_20250811_parrot_01_P5190526.drone_geolocated.csv; eventDate derived from flight identifier 20250811; eventTime recovered from the video container creation time (release build 2026-11-10; metadata/provenance/drone_time_recovery_20250811.json); coordinate_source=original_video_quicktime_location_static; not per-frame telemetry,09:19:55/09:23:43
+TheWilds_2025_drone_20250818_parrot_01_P5760583,TheWilds_2025_drone_20250818_parrot_01,MediaEvent,2025-08-18,Drone video P5760583; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,9832,classified detection crops,The Wilds Conservation Center,20250818_parrot_01,39.83062300000001,-81.7380009999996,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250818_parrot_01; video: P5760583; detections: 9832; source CSV: thewilds_20250804_20250828_20250818_parrot_01_P5760583.drone_geolocated.csv,09:14:06/09:19:53
+TheWilds_2025_drone_20250818_parrot_01_P5760584,TheWilds_2025_drone_20250818_parrot_01,MediaEvent,2025-08-18,Drone video P5760584; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,14302,classified detection crops,The Wilds Conservation Center,20250818_parrot_01,39.829163298809085,-81.74072732544536,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250818_parrot_01; video: P5760584; detections: 14302; source CSV: thewilds_20250804_20250828_20250818_parrot_01_P5760584.drone_geolocated.csv,09:19:54/09:25:32
+TheWilds_2025_drone_20250818_parrot_01_P5760585,TheWilds_2025_drone_20250818_parrot_01,MediaEvent,2025-08-18,Drone video P5760585; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,864,classified detection crops,The Wilds Conservation Center,20250818_parrot_01,39.82978903081466,-81.73963306787806,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250818_parrot_01; video: P5760585; detections: 864; source CSV: thewilds_20250804_20250828_20250818_parrot_01_P5760585.drone_geolocated.csv,09:25:32/09:26:47
+TheWilds_2025_drone_20250818_parrot_02_P5770586,TheWilds_2025_drone_20250818_parrot_02,MediaEvent,2025-08-18,Drone video P5770586; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,7569,classified detection crops,The Wilds Conservation Center,20250818_parrot_02,39.83553983333299,-81.729999,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250818_parrot_02; video: P5770586; detections: 7569; source CSV: thewilds_20250804_20250828_20250818_parrot_02_P5770586.drone_geolocated.csv,09:43:25/09:48:59
+TheWilds_2025_drone_20250818_parrot_02_P5770587,TheWilds_2025_drone_20250818_parrot_02,MediaEvent,2025-08-18,Drone video P5770587; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,6426,classified detection crops,The Wilds Conservation Center,20250818_parrot_02,39.83540453324248,-81.73063304563993,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250818_parrot_02; video: P5770587; detections: 6426; source CSV: thewilds_20250804_20250828_20250818_parrot_02_P5770587.drone_geolocated.csv,09:49:00/09:54:35
+TheWilds_2025_drone_20250818_parrot_02_P5770588,TheWilds_2025_drone_20250818_parrot_02,MediaEvent,2025-08-18,Drone video P5770588; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,2050,classified detection crops,The Wilds Conservation Center,20250818_parrot_02,39.83514552833084,-81.73167434590653,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250818_parrot_02; video: P5770588; detections: 2050; source CSV: thewilds_20250804_20250828_20250818_parrot_02_P5770588.drone_geolocated.csv,09:54:35/09:56:02
+TheWilds_2025_drone_20250825_parrot_01_P6140626,TheWilds_2025_drone_20250825_parrot_01,MediaEvent,2025-08-25,Drone video P6140626; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,9185,classified detection crops,The Wilds Conservation Center,20250825_parrot_01,39.8306344999969,-81.73801366666375,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250825_parrot_01; video: P6140626; detections: 9185; source CSV: thewilds_20250804_20250828_20250825_parrot_01_P6140626.drone_geolocated.csv,09:13:47/09:19:21
+TheWilds_2025_drone_20250825_parrot_01_P6140627,TheWilds_2025_drone_20250825_parrot_01,MediaEvent,2025-08-25,Drone video P6140627; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,3010,classified detection crops,The Wilds Conservation Center,20250825_parrot_01,39.829053662919705,-81.74227662655365,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250804_20250828; flight/session: 20250825_parrot_01; video: P6140627; detections: 3010; source CSV: thewilds_20250804_20250828_20250825_parrot_01_P6140627.drone_geolocated.csv,09:19:22/09:20:56
+TheWilds_2025_drone_20250908_djiair_1_DJI_0070,TheWilds_2025_drone_20250908_djiair_1,MediaEvent,2025-09-08,Drone video DJI_0070; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,14940,classified detection crops,The Wilds Conservation Center,20250908_djiair_1,39.8306117686427,-81.7381286438897,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250908_20250908; flight/session: 20250908_djiair_1; video: DJI_0070; detections: 14940; source CSV: thewilds_20250908_20250908_20250908_djiair_1_DJI_0070.drone_geolocated.csv,09:16:51/09:23:25
+TheWilds_2025_drone_20250908_djiair_1_DJI_0071,TheWilds_2025_drone_20250908_djiair_1,MediaEvent,2025-09-08,Drone video DJI_0071; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,5266,classified detection crops,The Wilds Conservation Center,20250908_djiair_1,39.8289907319909,-81.7419589340118,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250908_20250908; flight/session: 20250908_djiair_1; video: DJI_0071; detections: 5266; source CSV: thewilds_20250908_20250908_20250908_djiair_1_DJI_0071.drone_geolocated.csv,09:23:51/09:27:29
+TheWilds_2025_drone_20250908_djiair_2_DJI_0072,TheWilds_2025_drone_20250908_djiair_2,MediaEvent,2025-09-08,Drone video DJI_0072; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,5327,classified detection crops,The Wilds Conservation Center,20250908_djiair_2,39.8306056257256,-81.7381547387215,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250908_20250908; flight/session: 20250908_djiair_2; video: DJI_0072; detections: 5327; source CSV: thewilds_20250908_20250908_20250908_djiair_2_DJI_0072.drone_geolocated.csv,09:29:14/09:42:03
+TheWilds_2025_drone_20250908_djiair_3_DJI_0074,TheWilds_2025_drone_20250908_djiair_3,MediaEvent,2025-09-08,Drone video DJI_0074; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,567,classified detection crops,The Wilds Conservation Center,20250908_djiair_3,39.8341484494756,-81.7376652256038,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250908_20250908; flight/session: 20250908_djiair_3; video: DJI_0074; detections: 567; source CSV: thewilds_20250908_20250908_20250908_djiair_3_DJI_0074.drone_geolocated.csv,10:25:29/10:29:28
+TheWilds_2025_drone_20250908_djiair_4_DJI_0075,TheWilds_2025_drone_20250908_djiair_4,MediaEvent,2025-09-08,Drone video DJI_0075; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,53476,classified detection crops,The Wilds Conservation Center,20250908_djiair_4,39.834147453031,-81.7377155569067,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250908_20250908; flight/session: 20250908_djiair_4; video: DJI_0075; detections: 53476; source CSV: thewilds_20250908_20250908_20250908_djiair_4_DJI_0075.drone_geolocated.csv,10:30:05/10:43:44
+TheWilds_2025_drone_20250908_djiair_4_DJI_0076,TheWilds_2025_drone_20250908_djiair_4,MediaEvent,2025-09-08,Drone video DJI_0076; SAM3 detections with BioCLIP Tree-of-Life labels and projected animal positions where available,3140,classified detection crops,The Wilds Conservation Center,20250908_djiair_4,39.8329696799844,-81.734909052093,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250908_20250908; flight/session: 20250908_djiair_4; video: DJI_0076; detections: 3140; source CSV: thewilds_20250908_20250908_20250908_djiair_4_DJI_0076.drone_geolocated.csv,10:43:44/10:44:59
+TheWilds_2025_drone_20250703_parrotanafi_4_VID_001,TheWilds_2025_drone_20250703_parrotanafi_4,MediaEvent,2025-07-03,Drone video VID_001; intentionally excluded from SAM3 and Tree-of-Life processing after review found no significant content.,0,classified detection crops,The Wilds Conservation Center,20250703_parrotanafi_4,39.830678500000175,-81.73834333333349,100,WGS84,United States,Ohio,Muskingum,TheWilds_2025_drones,The Wilds 2025 Multi-Modal Survey - Drone Imagery,Drone field team,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_4; video: VID_001; processingStatus: intentionally_excluded; reason: tiny video reviewed with no significant content; raw media: thewilds_20250630_20250703/drones/20250703_parrotanafi_4/video_files/VID_001.MP4,

smartwilds-drones/media_manifest.csv

@@ -0,0 +1,78 @@
+release_path,source_path,bytes,include,exclusion_reason
+media/thewilds_20250630_20250703/20250701_parrotanafi_1/metadata/0914_2025-07-01T083438-0400_10D6BC_0018.json,thewilds_20250630_20250703/drones/20250701_parrotanafi_1/metadata/0914_2025-07-01T083438-0400_10D6BC_0018.json,1824672,true,
+media/thewilds_20250630_20250703/20250701_parrotanafi_1/metadata/File.jpg,thewilds_20250630_20250703/drones/20250701_parrotanafi_1/metadata/File.jpg,473161,true,
+media/thewilds_20250630_20250703/20250701_parrotanafi_1/metadata/mission_notes.txt,thewilds_20250630_20250703/drones/20250701_parrotanafi_1/metadata/mission_notes.txt,255,true,
+media/thewilds_20250630_20250703/20250701_parrotanafi_1/video_files/VID_001.MP4,thewilds_20250630_20250703/drones/20250701_parrotanafi_1/video_files/VID_001.MP4,109558823,true,
+media/thewilds_20250630_20250703/20250701_parrotanafi_1/video_files/VID_002.MP4,thewilds_20250630_20250703/drones/20250701_parrotanafi_1/video_files/VID_002.MP4,4205523945,true,
+media/thewilds_20250630_20250703/20250701_parrotanafi_1/video_files/VID_003.MP4,thewilds_20250630_20250703/drones/20250701_parrotanafi_1/video_files/VID_003.MP4,4210137951,true,
+media/thewilds_20250630_20250703/20250701_parrotanafi_1/video_files/VID_004.MP4,thewilds_20250630_20250703/drones/20250701_parrotanafi_1/video_files/VID_004.MP4,1432755297,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_1/metadata/0914_2025-07-03T154754-0400_16A458_0019.json,thewilds_20250630_20250703/drones/20250703_parrotanafi_1/metadata/0914_2025-07-03T154754-0400_16A458_0019.json,397407,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_1/metadata/0914_2025-07-03T155059-0400_54039C_001A.json,thewilds_20250630_20250703/drones/20250703_parrotanafi_1/metadata/0914_2025-07-03T155059-0400_54039C_001A.json,4135502,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_1/metadata/mission_notes_1.txt,thewilds_20250630_20250703/drones/20250703_parrotanafi_1/metadata/mission_notes_1.txt,115,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_1/video_files/VID_001.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_1/video_files/VID_001.MP4,2218716678,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_1/video_files/VID_002.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_1/video_files/VID_002.MP4,4206460839,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_1/video_files/VID_003.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_1/video_files/VID_003.MP4,1269341411,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_2/metadata/0914_2025-07-03T162104-0400_BF7B18_001D.json,thewilds_20250630_20250703/drones/20250703_parrotanafi_2/metadata/0914_2025-07-03T162104-0400_BF7B18_001D.json,3100196,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_2/metadata/mission_notes_2.txt,thewilds_20250630_20250703/drones/20250703_parrotanafi_2/metadata/mission_notes_2.txt,115,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_2/video_files/VID_001.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_2/video_files/VID_001.MP4,93444816,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_2/video_files/VID_002.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_2/video_files/VID_002.MP4,4208100144,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_2/video_files/VID_003.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_2/video_files/VID_003.MP4,3957176852,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_3/metadata/0914_2025-07-03T165620-0400_D602EB_001F.json,thewilds_20250630_20250703/drones/20250703_parrotanafi_3/metadata/0914_2025-07-03T165620-0400_D602EB_001F.json,1673349,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_3/metadata/mission_notes_3.txt,thewilds_20250630_20250703/drones/20250703_parrotanafi_3/metadata/mission_notes_3.txt,115,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_3/video_files/VID_001.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_3/video_files/VID_001.MP4,270997992,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_3/video_files/VID_002.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_3/video_files/VID_002.MP4,3802449394,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_4/metadata/0914_2025-07-03T172345-0400_75DB7A_0021.json,thewilds_20250630_20250703/drones/20250703_parrotanafi_4/metadata/0914_2025-07-03T172345-0400_75DB7A_0021.json,1195608,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_4/metadata/mission_notes_4.txt,thewilds_20250630_20250703/drones/20250703_parrotanafi_4/metadata/mission_notes_4.txt,101,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_4/video_files/VID_001.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_4/video_files/VID_001.MP4,13888970,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_4/video_files/VID_002.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_4/video_files/VID_002.MP4,4201193970,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_4/video_files/VID_003.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_4/video_files/VID_003.MP4,2441912991,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_5/metadata/0914_2025-07-03T172555-0400_0F0BAD_0008.json,thewilds_20250630_20250703/drones/20250703_parrotanafi_5/metadata/0914_2025-07-03T172555-0400_0F0BAD_0008.json,2670278,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_5/metadata/PNG image.png,thewilds_20250630_20250703/drones/20250703_parrotanafi_5/metadata/PNG image.png,8327847,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_5/metadata/mission_notes_5.txt,thewilds_20250630_20250703/drones/20250703_parrotanafi_5/metadata/mission_notes_5.txt,100,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_5/videos/VID_001.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_5/videos/VID_001.MP4,59665488,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_5/videos/VID_002.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_5/videos/VID_002.MP4,4207387557,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_5/videos/VID_003.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_5/videos/VID_003.MP4,2675309996,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_5/videos/VID_004.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_5/videos/VID_004.MP4,4208574991,true,
+media/thewilds_20250630_20250703/20250703_parrotanafi_5/videos/VID_005.MP4,thewilds_20250630_20250703/drones/20250703_parrotanafi_5/videos/VID_005.MP4,1495358825,true,
+media/thewilds_20250804_20250828/20250811_DJI_01/metadata/ExportCSV_2025-08-11_[10-29-13].csv,thewilds_20250804_20250828/drones/20250811_DJI_01/metadata/ExportCSV_2025-08-11_[10-29-13].csv,207909,true,
+media/thewilds_20250804_20250828/20250811_DJI_01/videos/DJI_0001.MOV,thewilds_20250804_20250828/drones/20250811_DJI_01/videos/DJI_0001.MOV,735322563,true,
+media/thewilds_20250804_20250828/20250811_DJI_01/videos/DJI_0002.MOV,thewilds_20250804_20250828/drones/20250811_DJI_01/videos/DJI_0002.MOV,1895502839,true,
+media/thewilds_20250804_20250828/20250811_parrot_01/videos/P5180524.MP4,thewilds_20250804_20250828/drones/20250811_parrot_01/videos/P5180524.MP4,1148007773,true,
+media/thewilds_20250804_20250828/20250811_parrot_01/videos/P5190525.MP4,thewilds_20250804_20250828/drones/20250811_parrot_01/videos/P5190525.MP4,4209204818,true,
+media/thewilds_20250804_20250828/20250811_parrot_01/videos/P5190526.MP4,thewilds_20250804_20250828/drones/20250811_parrot_01/videos/P5190526.MP4,2848054800,true,
+media/thewilds_20250804_20250828/20250818_parrot_01/metadata/0914_2025-08-18T091406-0400_3B8915_008A.json,thewilds_20250804_20250828/drones/20250818_parrot_01/metadata/0914_2025-08-18T091406-0400_3B8915_008A.json,1791153,true,
+media/thewilds_20250804_20250828/20250818_parrot_01/videos/P5760583.MP4,thewilds_20250804_20250828/drones/20250818_parrot_01/videos/P5760583.MP4,4208941916,true,
+media/thewilds_20250804_20250828/20250818_parrot_01/videos/P5760584.MP4,thewilds_20250804_20250828/drones/20250818_parrot_01/videos/P5760584.MP4,4212081400,true,
+media/thewilds_20250804_20250828/20250818_parrot_01/videos/P5760585.MP4,thewilds_20250804_20250828/drones/20250818_parrot_01/videos/P5760585.MP4,939266807,true,
+media/thewilds_20250804_20250828/20250818_parrot_02/metadata/0914_2025-08-18T094325-0400_43D0DB_008B.json,thewilds_20250804_20250828/drones/20250818_parrot_02/metadata/0914_2025-08-18T094325-0400_43D0DB_008B.json,1828935,true,
+media/thewilds_20250804_20250828/20250818_parrot_02/videos/P5770586.MP4,thewilds_20250804_20250828/drones/20250818_parrot_02/videos/P5770586.MP4,4207032621,true,
+media/thewilds_20250804_20250828/20250818_parrot_02/videos/P5770587.MP4,thewilds_20250804_20250828/drones/20250818_parrot_02/videos/P5770587.MP4,4207690420,true,
+media/thewilds_20250804_20250828/20250818_parrot_02/videos/P5770588.MP4,thewilds_20250804_20250828/drones/20250818_parrot_02/videos/P5770588.MP4,1093737528,true,
+media/thewilds_20250804_20250828/20250825_parrot_01/metadata/2025-08-25 09-13-47.json,thewilds_20250804_20250828/drones/20250825_parrot_01/metadata/2025-08-25 09-13-47.json,603534,true,
+media/thewilds_20250804_20250828/20250825_parrot_01/videos/P6140626.MP4,thewilds_20250804_20250828/drones/20250825_parrot_01/videos/P6140626.MP4,4208317405,true,
+media/thewilds_20250804_20250828/20250825_parrot_01/videos/P6140627.MP4,thewilds_20250804_20250828/drones/20250825_parrot_01/videos/P6140627.MP4,1182758133,true,
+media/thewilds_20250804_20250828/20250825_parrot_02/metadata/2025-08-25 09-43-01.json,thewilds_20250804_20250828/drones/20250825_parrot_02/metadata/2025-08-25 09-43-01.json,1289427,true,
+media/thewilds_20250804_20250828/20250825_parrot_03/metadata/2025-08-25 10-15-53.json,thewilds_20250804_20250828/drones/20250825_parrot_03/metadata/2025-08-25 10-15-53.json,1285174,true,
+media/thewilds_20250804_20250828/20250825_parrot_04/metadata/2025-08-25 10-50-31.json,thewilds_20250804_20250828/drones/20250825_parrot_04/metadata/2025-08-25 10-50-31.json,741319,true,
+media/thewilds_20250908_20250908/20250908_djiair_1/metadata/Sep-8th-2025-09-16AM-Flight-Airdata.csv,thewilds_20250908_20250908/drones/20250908_djiair_1/metadata/Sep-8th-2025-09-16AM-Flight-Airdata.csv,1724364,true,
+media/thewilds_20250908_20250908/20250908_djiair_1/metadata/Sep-8th-2025-09-16AM-Flight-Airdata.kml,thewilds_20250908_20250908/drones/20250908_djiair_1/metadata/Sep-8th-2025-09-16AM-Flight-Airdata.kml,46693,true,
+media/thewilds_20250908_20250908/20250908_djiair_1/metadata/airdata.png,thewilds_20250908_20250908/drones/20250908_djiair_1/metadata/airdata.png,1064278,true,
+media/thewilds_20250908_20250908/20250908_djiair_1/metadata/mission_notes.txt,thewilds_20250908_20250908/drones/20250908_djiair_1/metadata/mission_notes.txt,429,true,
+media/thewilds_20250908_20250908/20250908_djiair_1/video_files/DJI_0070.MP4,thewilds_20250908_20250908/drones/20250908_djiair_1/video_files/DJI_0070.MP4,1985030787,true,
+media/thewilds_20250908_20250908/20250908_djiair_1/video_files/DJI_0071.MP4,thewilds_20250908_20250908/drones/20250908_djiair_1/video_files/DJI_0071.MP4,1109454083,true,
+media/thewilds_20250908_20250908/20250908_djiair_2/metadata/Sep-8th-2025-09-29AM-Flight-Airdata.csv,thewilds_20250908_20250908/drones/20250908_djiair_2/metadata/Sep-8th-2025-09-29AM-Flight-Airdata.csv,1022440,true,
+media/thewilds_20250908_20250908/20250908_djiair_2/metadata/Sep-8th-2025-09-29AM-Flight-Airdata.kml,thewilds_20250908_20250908/drones/20250908_djiair_2/metadata/Sep-8th-2025-09-29AM-Flight-Airdata.kml,27987,true,
+media/thewilds_20250908_20250908/20250908_djiair_2/metadata/airdata2.png,thewilds_20250908_20250908/drones/20250908_djiair_2/metadata/airdata2.png,1058708,true,
+media/thewilds_20250908_20250908/20250908_djiair_2/metadata/mission_notes.txt,thewilds_20250908_20250908/drones/20250908_djiair_2/metadata/mission_notes.txt,417,true,
+media/thewilds_20250908_20250908/20250908_djiair_2/videos/DJI_0072.MP4,thewilds_20250908_20250908/drones/20250908_djiair_2/videos/DJI_0072.MP4,2279725035,true,
+media/thewilds_20250908_20250908/20250908_djiair_3/metadata/Sep-8th-2025-10-25AM-Flight-Airdata.csv,thewilds_20250908_20250908/drones/20250908_djiair_3/metadata/Sep-8th-2025-10-25AM-Flight-Airdata.csv,247412,true,
+media/thewilds_20250908_20250908/20250908_djiair_3/metadata/Sep-8th-2025-10-25AM-Flight-Airdata.kml,thewilds_20250908_20250908/drones/20250908_djiair_3/metadata/Sep-8th-2025-10-25AM-Flight-Airdata.kml,7964,true,
+media/thewilds_20250908_20250908/20250908_djiair_3/metadata/airdata3.png,thewilds_20250908_20250908/drones/20250908_djiair_3/metadata/airdata3.png,713462,true,
+media/thewilds_20250908_20250908/20250908_djiair_3/metadata/mission_notes.txt,thewilds_20250908_20250908/drones/20250908_djiair_3/metadata/mission_notes.txt,406,true,
+media/thewilds_20250908_20250908/20250908_djiair_3/videos/DJI_0074.MP4,thewilds_20250908_20250908/drones/20250908_djiair_3/videos/DJI_0074.MP4,1080092443,true,
+media/thewilds_20250908_20250908/20250908_djiair_4/metadata/Sep-8th-2025-10-30AM-Flight-Airdata.csv,thewilds_20250908_20250908/drones/20250908_djiair_4/metadata/Sep-8th-2025-10-30AM-Flight-Airdata.csv,3591271,true,
+media/thewilds_20250908_20250908/20250908_djiair_4/metadata/Sep-8th-2025-10-30AM-Flight-Airdata.kml,thewilds_20250908_20250908/drones/20250908_djiair_4/metadata/Sep-8th-2025-10-30AM-Flight-Airdata.kml,95748,true,
+media/thewilds_20250908_20250908/20250908_djiair_4/metadata/airdata4.png,thewilds_20250908_20250908/drones/20250908_djiair_4/metadata/airdata4.png,968634,true,
+media/thewilds_20250908_20250908/20250908_djiair_4/metadata/mission_notes.txt,thewilds_20250908_20250908/drones/20250908_djiair_4/metadata/mission_notes.txt,276,true,
+media/thewilds_20250908_20250908/20250908_djiair_4/videos/DJI_0075.MP4,thewilds_20250908_20250908/drones/20250908_djiair_4/videos/DJI_0075.MP4,4090477321,true,
+media/thewilds_20250908_20250908/20250908_djiair_4/videos/DJI_0076.MP4,thewilds_20250908_20250908/drones/20250908_djiair_4/videos/DJI_0076.MP4,386409656,true,

smartwilds-drones/metadata/DATA_DICTIONARY.md

@@ -0,0 +1,181 @@
+# smartwilds-drones data dictionary
+
+Generated by `scripts/build_hf_release.py`. Darwin Core term definitions: <https://dwc.tdwg.org/terms/>. Columns not listed in a Darwin Core namespace are SmartWilds-specific.
+
+## `analysis/bird_water_association_per_box.csv`
+
+1,590 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `modality` |  | camera_trap |
+| `session` |  | camera_trap_curated_birds |
+| `image_id` |  | 1 |
+| `annotation_id` |  | 1 |
+| `species_label` |  | canada_goose |
+| `file_name` |  | images/canada_goose/001_canada_goose_TW09_CT02_001_camera_tr… |
+| `bbox` |  | [2120.788818359375, 1105.6610107421875, 123.6484375, 58.1750… |
+| `box_area_pixels` |  | 7193.253889083862 |
+| `water_pixels_in_box` |  | 149 |
+| `water_overlap_fraction` |  | 1.0 |
+| `has_any_water_overlap` |  | True |
+| `water_rule` |  | camera provenance: any of ['TW09'] in file_name => water; ot… |
+| `lake_json` |  | TheWilds2025/processing_output/detections/drones/drone_every… |
+| `n_lake_detections` |  | 1 |
+| `max_lake_score` |  | 0.9739968180656433 |
+| `water_overlap_ge_0.01` |  | True |
+| `water_overlap_ge_0.05` |  | True |
+| `water_overlap_ge_0.1` |  | True |
+| `water_overlap_ge_0.25` |  | True |
+| `water_overlap_ge_0.5` |  | True |
+
+## `data/detections.csv`
+
+317,559 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `occurrenceID` | Darwin Core: unique Occurrence identifier | TheWilds_2025_drone_000001 |
+| `eventID` | Darwin Core: unique Event identifier | TheWilds_2025_drone_20250701_parrotanafi_1_VID_001 |
+| `video_path` | Release-relative path of the raw video | media/thewilds_20250630_20250703/20250701_parrotanafi_1/vide… |
+| `detection_id` | <video>_<frame>_<k> from the SAM3 run | VID_001_000105_001 |
+| `frame_index` | 0-based native frame index in the video (SAM3 sampled every 15th frame) | 105 |
+| `bbox_x` | Box left, pixels, in the native frame | 8.42 |
+| `bbox_y` | Box top, pixels | -8.39 |
+| `bbox_w` | Box width, pixels | 74.68 |
+| `bbox_h` | Box height, pixels | 68.38 |
+| `scientificName` | Darwin Core: top-1 model or annotator taxon | Kryptopterus vitreolus |
+| `vernacularName` | Darwin Core: common name | Common Starling |
+| `top1_score` | BioCLIP Tree-of-Life top-1 score | 0.0019780425354838 |
+| `georef_status` | Ground-projection status (ok | invalid_altitude | missing_drone_gps | horizon_ray | out_of_range) | invalid_altitude |
+
+## `data/events.csv`
+
+55 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `eventID` | Darwin Core: unique Event identifier | TheWilds_2025_drones |
+| `parentEventID` | Darwin Core: eventID of the parent Event | TheWilds_2025 |
+| `eventType` | Darwin Core: Project, SensorDeployment, AudioRecording, MediaEvent, or ObservationSession | Project |
+| `eventDate` | Darwin Core: ISO 8601 date or date interval (start/end) | 2025-07-01/2025-09-08 |
+| `samplingProtocol` | Darwin Core: sensor, schedule, and processing chain | Drone video surveys; detections by SAM3, Tree-of-Life classi… |
+| `sampleSizeValue` | Darwin Core: sample size | 39 |
+| `sampleSizeUnit` | Darwin Core: unit of sampleSizeValue | raw drone videos (38 processed; 1 intentionally excluded) |
+| `locality` | Darwin Core: named place | The Wilds Conservation Center, Muskingum County, Ohio, USA |
+| `locationID` | Darwin Core: site (TWxx) for fixed sensors; flight ID for drones | 20250701_parrotanafi_1 |
+| `decimalLatitude` | Darwin Core: WGS84; see dataGeneralizations for what the point represents | 39.83 |
+| `decimalLongitude` | Darwin Core: WGS84 | -81.74 |
+| `coordinateUncertaintyInMeters` | Darwin Core: meters | 5000 |
+| `geodeticDatum` | Darwin Core: WGS84 | WGS84 |
+| `country` |  | United States |
+| `stateProvince` |  | Ohio |
+| `county` |  | Muskingum |
+| `datasetID` | Darwin Core: modality dataset identifier | TheWilds_2025_drones |
+| `datasetName` | Darwin Core | The Wilds 2025 Multi-Modal Survey - Drone Imagery |
+| `recordedBy` | Darwin Core: device or team | Drone field team |
+| `fieldNotes` | Darwin Core: provenance and field notes | 38 videos regenerated from current Tree-of-Life outputs; VID… |
+| `eventTime` | Darwin Core: local time (or interval); blank when not recoverable | 10:29:39/10:35:44 |
+
+## `data/occurrences.csv`
+
+318,215 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `occurrenceID` | Darwin Core: unique Occurrence identifier | TheWilds_2025_drone_000001 |
+| `eventID` | Darwin Core: unique Event identifier | TheWilds_2025_drone_20250701_parrotanafi_1_VID_001 |
+| `catalogNumber` | Darwin Core: source record identifier | drone_20250701_parrotanafi_1_VID_001_image1_ann1 |
+| `scientificName` | Darwin Core: top-1 model or annotator taxon | Kryptopterus vitreolus |
+| `kingdom` | Darwin Core: rank | Animalia |
+| `phylum` | Darwin Core: rank | Chordata |
+| `class` | Darwin Core: rank | Aves |
+| `order` | Darwin Core: rank | Siluriformes |
+| `family` | Darwin Core: rank | Siluridae |
+| `genus` | Darwin Core: rank | Kryptopterus |
+| `specificEpithet` | Darwin Core: rank | vitreolus |
+| `taxonRank` | Darwin Core: rank of scientificName | species |
+| `vernacularName` | Darwin Core: common name | Common Starling |
+| `occurrenceStatus` | Darwin Core: present | present |
+| `individualCount` | Darwin Core: 1 per visual crop; number of BirdNET detections for acoustic rows | 1 |
+| `lifeStage` |  |  |
+| `sex` |  |  |
+| `behavior` |  |  |
+| `basisOfRecord` | Darwin Core: MachineObservation or HumanObservation | MachineObservation |
+| `recordedBy` | Darwin Core: device or team | Drone imagery |
+| `identifiedBy` | Darwin Core: model (and version) or annotator | SAM3 + BioCLIP Tree of Life |
+| `identificationVerificationStatus` | Darwin Core: verification status; the vocabulary differs by table (SUPPLEMENTAL.md S9.3) | unverified |
+| `identificationRemarks` | Darwin Core: rank, top-1 score, bbox xywh (px), detector label | Tree-of-Life rank: species; top1 score: 0.001978042535483837… |
+| `eventDate` | Darwin Core: ISO 8601 date or date interval (start/end) | 2025-07-01 |
+| `eventTime` | Darwin Core: local time (or interval); blank when not recoverable | 08:34:41 |
+| `decimalLatitude` | Darwin Core: WGS84; see dataGeneralizations for what the point represents | 39.82887083333163 |
+| `decimalLongitude` | Darwin Core: WGS84 | -81.74238066666636 |
+| `coordinateUncertaintyInMeters` | Darwin Core: meters | 100 |
+| `geodeticDatum` | Darwin Core: WGS84 | WGS84 |
+| `locality` | Darwin Core: named place | The Wilds Conservation Center |
+| `locationID` | Darwin Core: site (TWxx) for fixed sensors; flight ID for drones | 20250701_parrotanafi_1 |
+| `country` |  | United States |
+| `stateProvince` |  | Ohio |
+| `county` |  | Muskingum |
+| `establishmentMeans` |  | managed |
+| `degreeOfEstablishment` |  | captive |
+| `occurrenceRemarks` | Darwin Core: projection status, timestamp method, repair notes (key=value; ...) | Frame-level drone detection crop; source=deep learning (sam3… |
+| `associatedMedia` | Darwin Core: release-relative path to the source media file | media/thewilds_20250630_20250703/20250701_parrotanafi_1/vide… |
+| `associatedReferences` | Darwin Core: release-relative supporting analysis file |  |
+| `datasetID` | Darwin Core: modality dataset identifier | TheWilds_2025_drones |
+| `datasetName` | Darwin Core | The Wilds 2025 Multi-Modal Survey - Drone Imagery |
+| `informationWithheld` | Darwin Core |  |
+| `dataGeneralizations` | Darwin Core: what the coordinate represents | When drone_georef_status=ok, decimalLatitude/decimalLongitud… |
+
+## `media_manifest.csv`
+
+77 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `release_path` | Path of the file inside this dataset | media/thewilds_20250630_20250703/20250701_parrotanafi_1/meta… |
+| `source_path` | Path in the project archive | thewilds_20250630_20250703/drones/20250701_parrotanafi_1/met… |
+| `bytes` | File size | 1824672 |
+| `include` | true if the file is distributed in media/ | true |
+| `exclusion_reason` | Why a file is not distributed |  |
+
+## `metadata/drone_projection_config.csv`
+
+4 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `enabled` |  | 1 |
+| `pattern` |  | parrotanafi |
+| `hfov_deg` |  | 69.0 |
+| `vfov_deg` |  | 42.5 |
+| `fallback_pitch_deg` |  | -45.0 |
+| `use_fallback_when_pitch_abs_lt` |  | 35.0 |
+| `notes` |  | Parrot Anafi approximate 4K FOV; processed gimbal_pitch ofte… |
+
+## `metadata/flights.csv`
+
+15 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `eventID` | Darwin Core: unique Event identifier | TheWilds_2025_drone_20250701_parrotanafi_1 |
+| `eventDate` | Darwin Core: ISO 8601 date or date interval (start/end) | 2025-07-01 |
+| `locationID` | Darwin Core: site (TWxx) for fixed sensors; flight ID for drones | 20250701_parrotanafi_1 |
+| `fieldNotes` | Darwin Core: provenance and field notes | Collection: thewilds_20250630_20250703; flight/session: 2025… |
+| `platform` |  | Parrot Anafi 4K |
+| `videos` |  | 4 |
+| `occurrences` |  | 16528 |
+
+## `metadata/platforms.csv`
+
+3 rows.
+
+| Column | Description | Example |
+|---|---|---|
+| `Platform` |  | Parrot Anafi 4K |
+| `Campaigns` |  | C1 (Jul 1, 3), C2 (Aug 11, 18, 25) |
+| `Video Resolution` |  | 4K MP4 |
+| `Telemetry Format` |  | Parrot JSON / GUTMA DX |
+| `Flights` |  | 10 video-bearing sessions |
+

smartwilds-drones/metadata/drone_projection_config.csv

@@ -0,0 +1,5 @@
+enabled,pattern,hfov_deg,vfov_deg,fallback_pitch_deg,use_fallback_when_pitch_abs_lt,notes
+1,parrotanafi,69.0,42.5,-45.0,35.0,"Parrot Anafi approximate 4K FOV; processed gimbal_pitch often resembles body pitch, so near-level values use -45 deg fallback."
+1,parrot,69.0,42.5,-45.0,35.0,"Parrot Anafi approximate 4K FOV; processed gimbal_pitch often resembles body pitch, so near-level values use -45 deg fallback."
+1,djiair,84.0,53.1,-45.0,2.0,"DJI Air approximate wide video FOV; use gimbal_pitch unless effectively zero."
+1,DJI_01,84.0,53.1,-45.0,2.0,"DJI August flight approximate video FOV; projection will fail if pose metadata is absent."

smartwilds-drones/metadata/flights.csv

@@ -0,0 +1,16 @@
+eventID,eventDate,locationID,fieldNotes,platform,videos,occurrences
+TheWilds_2025_drone_20250701_parrotanafi_1,2025-07-01,20250701_parrotanafi_1,Collection: thewilds_20250630_20250703; flight/session: 20250701_parrotanafi_1,Parrot Anafi 4K,4,16528
+TheWilds_2025_drone_20250703_parrotanafi_1,2025-07-03,20250703_parrotanafi_1,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_1,Parrot Anafi 4K,3,42098
+TheWilds_2025_drone_20250703_parrotanafi_2,2025-07-03,20250703_parrotanafi_2,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_2,Parrot Anafi 4K,3,9822
+TheWilds_2025_drone_20250703_parrotanafi_3,2025-07-03,20250703_parrotanafi_3,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_3,Parrot Anafi 4K,2,692
+TheWilds_2025_drone_20250703_parrotanafi_4,2025-07-03,20250703_parrotanafi_4,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_4,Parrot Anafi 4K,3,7879
+TheWilds_2025_drone_20250703_parrotanafi_5,2025-07-03,20250703_parrotanafi_5,Collection: thewilds_20250630_20250703; flight/session: 20250703_parrotanafi_5; coordinate_source=unavailable_no_single_session_location,Parrot Anafi 4K,5,63718
+TheWilds_2025_drone_20250811_DJI_01,2025-08-11,20250811_DJI_01,Collection: thewilds_20250804_20250828; flight/session: 20250811_DJI_01; eventDate derived from flight identifier 20250811; eventTime recovered from the video container creation time (release build 2026-11-10; metadata/provenance/drone_time_recovery_20250811.json),DJI (unspecified airframe),2,17057
+TheWilds_2025_drone_20250811_parrot_01,2025-08-11,20250811_parrot_01,Collection: thewilds_20250804_20250828; flight/session: 20250811_parrot_01; eventDate derived from flight identifier 20250811; eventTime recovered from the video container creation time (release build 2026-11-10; metadata/provenance/drone_time_recovery_20250811.json); coordinate_source=unavailable_no_single_session_location,Parrot Anafi 4K,3,24467
+TheWilds_2025_drone_20250818_parrot_01,2025-08-18,20250818_parrot_01,Collection: thewilds_20250804_20250828; flight/session: 20250818_parrot_01,Parrot Anafi 4K,3,24998
+TheWilds_2025_drone_20250818_parrot_02,2025-08-18,20250818_parrot_02,Collection: thewilds_20250804_20250828; flight/session: 20250818_parrot_02,Parrot Anafi 4K,3,16045
+TheWilds_2025_drone_20250825_parrot_01,2025-08-25,20250825_parrot_01,Collection: thewilds_20250804_20250828; flight/session: 20250825_parrot_01,Parrot Anafi 4K,2,12195
+TheWilds_2025_drone_20250908_djiair_1,2025-09-08,20250908_djiair_1,Collection: thewilds_20250908_20250908; flight/session: 20250908_djiair_1,DJI Air 2,2,20206
+TheWilds_2025_drone_20250908_djiair_2,2025-09-08,20250908_djiair_2,Collection: thewilds_20250908_20250908; flight/session: 20250908_djiair_2,DJI Air 2,1,5327
+TheWilds_2025_drone_20250908_djiair_3,2025-09-08,20250908_djiair_3,Collection: thewilds_20250908_20250908; flight/session: 20250908_djiair_3,DJI Air 2,1,567
+TheWilds_2025_drone_20250908_djiair_4,2025-09-08,20250908_djiair_4,Collection: thewilds_20250908_20250908; flight/session: 20250908_djiair_4,DJI Air 2,2,56616

smartwilds-drones/metadata/platforms.csv

@@ -0,0 +1,4 @@
+Platform,Campaigns,Video Resolution,Telemetry Format,Flights
+Parrot Anafi 4K,"C1 (Jul 1, 3), C2 (Aug 11, 18, 25)",4K MP4,Parrot JSON / GUTMA DX,10 video-bearing sessions
+DJI (unspecified),C2 (Aug 11),1080p MOV,"CSV (GPS + battery, 10Hz)",1 session
+DJI Air 2,C3 (Sep 8),4K MP4,Airdata CSV,4 flights

smartwilds-drones/metadata/provenance/drone_missing_media_repair_20261010.json

@@ -0,0 +1,21 @@
+{
+  "policy": "Occurrences retain an extracted frame when available; unavailable scratch frames are represented by their unique retained original video and explicitly marked as a video-level media fallback.",
+  "previous_audit": {
+    "unique_missing_scratch_frame_paths": 16069
+  },
+  "repair": "missing drone scratch-frame media references",
+  "tables": {
+    "drones/darwin_core_occurrences_drones.csv": {
+      "distinct_original_videos": 38,
+      "nonexistent_media_after_repair": 0,
+      "rows_relinked_to_original_video": 307260,
+      "rows_retaining_existing_scratch_frame": 10943
+    },
+    "occurrences_all_modalities.csv": {
+      "distinct_original_videos": 38,
+      "nonexistent_media_after_repair": 0,
+      "rows_relinked_to_original_video": 307260,
+      "rows_retaining_existing_scratch_frame": 10943
+    }
+  }
+}

smartwilds-drones/metadata/provenance/drone_time_recovery_20250811.json

@@ -0,0 +1,116 @@
+{
+  "defect": 15,
+  "title": "Time of day of the 2025-08-11 drone flights recovered from video container creation times, with the clock convention resolved against reference flights and the field log",
+  "generated": "2026-11-10T20:20:49+00:00",
+  "generated_by": "release/scripts/recover_drone_times.py",
+  "timezone": "America/New_York (EDT, UTC-4) on 2025-08-11",
+  "method": "eventTime of each MediaEvent = container creation time (converted per airframe convention) to creation time + duration; eventTime of each Occurrence = video start + frame_index / fps; flight Events span their videos.",
+  "reference_flights": [
+    {
+      "airframe": "Parrot Anafi 4K",
+      "video": "P5760583.MP4",
+      "container_creation_time": "2025-08-18T13:14:08.000000Z",
+      "known_local_start": "2025-08-18T09:14:06-04:00",
+      "known_from": "metadata JSON 0914_2025-08-18T091406-0400_3B8915_008A.json",
+      "offset_if_container_is_utc_s": 2.0,
+      "offset_if_container_is_local_s": 14402.0,
+      "convention": "UTC"
+    },
+    {
+      "airframe": "DJI Air 2",
+      "video": "DJI_0070.MP4",
+      "container_creation_time": "2025-09-08T13:16:47.000000Z",
+      "known_local_start": "2025-09-08T09:16:25-04:00",
+      "known_from": "telemetry datetime(utc) 2025-09-08 13:16:25 (first row)",
+      "offset_if_container_is_utc_s": 22.0,
+      "offset_if_container_is_local_s": 14422.0,
+      "convention": "UTC"
+    }
+  ],
+  "field_evidence": {
+    "field_log": "2025-08-11: 'flew 2 drones: parrot anafi near nomad shelter and zipline tower, DJI Mavic 3 at dock area'; Stop 1 (Nomad Ridge, TW01/TW02): 'Flew Parrot drone to survey lake area, captured images of Pdeer in lake'; the dock (TW09) was Stop 5.",
+    "jotform_checkins_local": [
+      "09:32 TW07 dropoff",
+      "09:32 TW01 pickup",
+      "09:40 TW07",
+      "09:41 TW06",
+      "09:44 TW07",
+      "09:53 TW07",
+      "09:55 TW06",
+      "10:48 TW05"
+    ],
+    "inference": "The team was on site from about 09:30. A Parrot start of 13:12 UTC is 09:12 local, consistent with flying at Stop 1 before servicing the TW01 camera at 09:32. A DJI start of 10:29 cannot be UTC (06:29 local, before anyone was on site), so the Mavic 3 wrote local time; 10:29 to 10:36 local fits the dock stop in the second half of the morning."
+  },
+  "videos": [
+    {
+      "eventID": "TheWilds_2025_drone_20250811_DJI_01_DJI_0001",
+      "flight_eventID": "TheWilds_2025_drone_20250811_DJI_01",
+      "video": "DJI_0001.MOV",
+      "airframe": "DJI Mavic 3",
+      "container_creation_time": "2025-08-11T10:29:39.000000Z",
+      "clock_convention": "local",
+      "local_start": "10:29:39",
+      "local_end": "10:31:17",
+      "duration_s": 98.098,
+      "fps": 29.97002997002997,
+      "frames": 2940,
+      "confidence": "medium: local-time convention inferred from the field schedule and the telemetry export file name; the Air 2 flown on 2025-09-08 wrote UTC, so the convention is airframe-specific"
+    },
+    {
+      "eventID": "TheWilds_2025_drone_20250811_DJI_01_DJI_0002",
+      "flight_eventID": "TheWilds_2025_drone_20250811_DJI_01",
+      "video": "DJI_0002.MOV",
+      "airframe": "DJI Mavic 3",
+      "container_creation_time": "2025-08-11T10:31:32.000000Z",
+      "clock_convention": "local",
+      "local_start": "10:31:32",
+      "local_end": "10:35:44",
+      "duration_s": 252.886,
+      "fps": 29.97002997002997,
+      "frames": 7579,
+      "confidence": "medium: local-time convention inferred from the field schedule and the telemetry export file name; the Air 2 flown on 2025-09-08 wrote UTC, so the convention is airframe-specific"
+    },
+    {
+      "eventID": "TheWilds_2025_drone_20250811_parrot_01_P5180524",
+      "flight_eventID": "TheWilds_2025_drone_20250811_parrot_01",
+      "video": "P5180524.MP4",
+      "airframe": "Parrot Anafi 4K",
+      "container_creation_time": "2025-08-11T13:12:43.000000Z",
+      "clock_convention": "UTC",
+      "local_start": "09:12:43",
+      "local_end": "09:14:14",
+      "duration_s": 91.264,
+      "fps": 29.980991941489382,
+      "frames": 2735,
+      "confidence": "high: UTC convention verified on the 2025-08-18 flight to within 2 s"
+    },
+    {
+      "eventID": "TheWilds_2025_drone_20250811_parrot_01_P5190525",
+      "flight_eventID": "TheWilds_2025_drone_20250811_parrot_01",
+      "video": "P5190525.MP4",
+      "airframe": "Parrot Anafi 4K",
+      "container_creation_time": "2025-08-11T13:14:19.000000Z",
+      "clock_convention": "UTC",
+      "local_start": "09:14:19",
+      "local_end": "09:19:54",
+      "duration_s": 335.339,
+      "fps": 29.973012359319455,
+      "frames": 10050,
+      "confidence": "high: UTC convention verified on the 2025-08-18 flight to within 2 s"
+    },
+    {
+      "eventID": "TheWilds_2025_drone_20250811_parrot_01_P5190526",
+      "flight_eventID": "TheWilds_2025_drone_20250811_parrot_01",
+      "video": "P5190526.MP4",
+      "airframe": "Parrot Anafi 4K",
+      "container_creation_time": "2025-08-11T13:19:55.000000Z",
+      "clock_convention": "UTC",
+      "local_start": "09:19:55",
+      "local_end": "09:23:43",
+      "duration_s": 228.27,
+      "fps": 29.974407175513495,
+      "frames": 6841,
+      "confidence": "high: UTC convention verified on the 2025-08-18 flight to within 2 s"
+    }
+  ]
+}

smartwilds-drones/metadata/provenance/excluded_drone_media_event_repair_20261008.json

@@ -0,0 +1,9 @@
+{
+  "repair": "explicit excluded raw drone MediaEvent",
+  "eventID": "TheWilds_2025_drone_20250703_parrotanafi_4_VID_001",
+  "raw_media": "thewilds_20250630_20250703/drones/20250703_parrotanafi_4/video_files/VID_001.MP4",
+  "processing_status": "intentionally_excluded",
+  "reason": "Tiny video reviewed with no significant content.",
+  "occurrences_added": 0,
+  "merged_event_count_after_repair": 2470
+}