Blame for croissant.jsonld

{
 "@context": {
  "@vocab": "https://schema.org/",
  "@language": "en",
  "sc": "https://schema.org/",
  "cr": "http://mlcommons.org/croissant/",
  "rai": "http://mlcommons.org/croissant/RAI/",
  "dct": "http://purl.org/dc/terms/",
  "citeAs": "cr:citeAs",
  "column": "cr:column",
  "conformsTo": "dct:conformsTo",
  "data": {
   "@id": "cr:data",
   "@type": "@json"
  },
  "dataType": {
   "@id": "cr:dataType",
   "@type": "@vocab"
  },
  "extract": "cr:extract",
  "field": "cr:field",
  "fileObject": "cr:fileObject",
  "fileSet": "cr:fileSet",
  "format": "cr:format",
  "includes": "cr:includes",
  "jsonPath": "cr:jsonPath",
  "key": "cr:key",
  "parentField": "cr:parentField",
  "path": "cr:path",
  "recordSet": "cr:recordSet",
  "references": "cr:references",
  "regex": "cr:regex",
  "repeated": "cr:repeated",
  "replace": "cr:replace",
  "separator": "cr:separator",
  "source": "cr:source",
  "subField": "cr:subField",
  "transform": "cr:transform",
  "prov": "http://www.w3.org/ns/prov#"
 },
 "@type": "sc:Dataset",
 "name": "WildBox",
 "alternateName": [
  "WildBox"
 ],
 "description": "WildBox: a 6-class monocular 3D wildlife detection benchmark. Drone footage of African megafauna (giraffe, plains zebra, Grevy's zebra, elephant, rhino, gazelle) with KITTI / Omni3D-compatible 3D bounding-box annotations. 45,979 train + 13,779 val frames at 1920x1080, across 64 per-video segments organised into 11 species-site groups. Annotations are released in two redundant forms: strict-KITTI per-frame txt files (lossy yaw-only rotation) and full-precision Omni3D / tracking-summary JSON (3x3 rotation matrix, centroid location). Release 1.1 exports the human-corrected annotation tracks for both label channels, ships the KITTI labels in a per-frame upright rectified frame with per-frame calibration files, and adds per-segment/per-video/per-track metadata CSVs (herd density and track inventory).",
 "url": "https://huggingface.co/datasets/3DOMFBK/WildBox",
 "license": "https://creativecommons.org/licenses/by/4.0/",
 "conformsTo": "http://mlcommons.org/croissant/1.1",
 "version": "1.1.0",
 "datePublished": "2026-10-29",
 "citeAs": "@inproceedings{shukla2026wildbox, title={{WildBox}: A Dataset and Benchmark for Aerial Monocular {3D} Detection of {African} Savanna Wildlife}, author={Shukla, Vandita and Meier, Kilian and Laporte-Devylder, Lucie and Rondeau Saint-Jean, Camille and Kline, Jenna M. and Costelloe, Blair R. and Tuia, Devis and Remondino, Fabio and Risse, Benjamin}, booktitle={Advances in Neural Information Processing Systems (NeurIPS), Evaluations and Datasets Track}, year={2026}, url={https://openreview.net/forum?id=AXawAJRwDi}, eprint={2606.21309}, archivePrefix={arXiv}, primaryClass={cs.CV}}",
 "creator": [
  {
   "@type": "sc:Person",
   "name": "Vandita Shukla"
  },
  {
   "@type": "sc:Person",
   "name": "Kilian Meier"
  },
  {
   "@type": "sc:Person",
   "name": "Lucie Laporte-Devylder"
  },
  {
   "@type": "sc:Person",
   "name": "Camille Rondeau Saint-Jean"
  },
  {
   "@type": "sc:Person",
   "name": "Jenna M. Kline"
  },
  {
   "@type": "sc:Person",
   "name": "Blair R. Costelloe"
  },
  {
   "@type": "sc:Person",
   "name": "Devis Tuia"
  },
  {
   "@type": "sc:Person",
   "name": "Fabio Remondino"
  },
  {
   "@type": "sc:Person",
   "name": "Benjamin Risse"
  }
 ],
 "keywords": [
  "monocular 3D detection",
  "wildlife",
  "open-vocabulary detection",
  "KITTI format",
  "Omni3D",
  "drone imagery"
 ],
 "distribution": [
  {
   "@type": "cr:FileObject",
   "@id": ".gitattributes",
   "name": ".gitattributes",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/.gitattributes",
   "encodingFormat": "text/plain",
   "contentSize": "2612",
   "sha256": "d65fdb298598f4bfb6f298fdac0d859b733f5a084a8278ca28d8b69da6e52fed"
  },
  {
   "@type": "cr:FileObject",
   "@id": "README.md",
   "name": "README.md",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/README.md",
   "encodingFormat": "text/markdown",
   "contentSize": "8424",
   "sha256": "7ac0acdd762b86729a2b2c4fb73cd3c8a2054a3014c270fea18c666e9e3c8cea"
  },
  {
   "@type": "cr:FileObject",
   "@id": "WildBox_train.json",
   "name": "WildBox_train.json",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/WildBox_train.json",
   "encodingFormat": "application/json",
   "contentSize": "241267367",
   "sha256": "e71d9b66d2e9bf97e84447c5634dbc5c22a9fbc1de389159f2c651ac28e2ae51"
  },
  {
   "@type": "cr:FileObject",
   "@id": "WildBox_val.json",
   "name": "WildBox_val.json",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/WildBox_val.json",
   "encodingFormat": "application/json",
   "contentSize": "93749408",
   "sha256": "f74ea346d1cbdf5837b62600b75c8e77ac69c28c4a6cdd8f27c3607961d9992e"
  },
  {
   "@type": "cr:FileObject",
   "@id": "checkpoints__CHECKPOINTS_README.md",
   "name": "CHECKPOINTS_README.md",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/checkpoints/CHECKPOINTS_README.md",
   "encodingFormat": "text/markdown",
   "contentSize": "3058",
   "sha256": "f32f05fad8f57d482703dbbfc8ec38f70092559643d4844897a6a218d6805e44"
  },
  {
   "@type": "cr:FileObject",
   "@id": "checkpoints__detany3d_ep2_seed0__checkpoint_0.pth",
   "name": "checkpoint_0.pth",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/checkpoints/detany3d_ep2_seed0/checkpoint_0.pth",
   "encodingFormat": "application/octet-stream",
   "contentSize": "4395140579",
   "sha256": "6d1b80cdfe88b5245c05c976667a6714809517742e7cefe27eeb40d7dfeed55a"
  },
  {
   "@type": "cr:FileObject",
   "@id": "checkpoints__ovmono3d_lift_init5sp_seed0__model_final.pth",
   "name": "model_final.pth",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/checkpoints/ovmono3d_lift_init5sp_seed0/model_final.pth",
   "encodingFormat": "application/octet-stream",
   "contentSize": "601426000",
   "sha256": "d80cfb9a394d848ec0bbe74c5e7f8c294c443a521ab94a590c6d84594204efd8"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep1__DJI_20240119151048_0001_V.zip",
   "name": "DJI_20240119151048_0001_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep1/DJI_20240119151048_0001_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "110912775",
   "sha256": "da5326f6c94cf8c0fa48eef49253d1bd988506e25cedf627ceab2b1e7a5d7f1f"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep1__DJI_20240119151437_0002_V.zip",
   "name": "DJI_20240119151437_0002_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep1/DJI_20240119151437_0002_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "315507146",
   "sha256": "7b34b67f7e1187b333a12ff4fc842d95a8b166f155f17382023b39c8564186fc"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep1__DJI_20240119151557_0003_V.zip",
   "name": "DJI_20240119151557_0003_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep1/DJI_20240119151557_0003_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "667994681",
   "sha256": "046f36c6ab12891255fee196ed9112783ac45f11c06022d135af39a8ad8533e5"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep1__DJI_20240119151946_0004_V.zip",
   "name": "DJI_20240119151946_0004_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep1/DJI_20240119151946_0004_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "1048661549",
   "sha256": "b0480d324951066e23dc8b3e7fe42f977a1ea3408c972ffa1704dd128c50791f"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep1__DJI_20240119152335_0005_V.zip",
   "name": "DJI_20240119152335_0005_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep1/DJI_20240119152335_0005_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "257779140",
   "sha256": "3cefcfe38ee0e1b8117f4d642b546a1ad9d710b46943843a09878da968c0f95f"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep1__DJI_20240119152552_0006_V.zip",
   "name": "DJI_20240119152552_0006_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep1/DJI_20240119152552_0006_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "231604273",
   "sha256": "0f9619fc9a0200abd22953ddc33b6386445a4f21f38a6794f24a31d500e3e216"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep2__DJI_20240626142859_0001_V.zip",
   "name": "DJI_20240626142859_0001_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep2/DJI_20240626142859_0001_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "623556657",
   "sha256": "1e510b5aef599a21eb047e4c8564e0c6211805a0346717b6f2f34f700d0d5e1b"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep2__DJI_20240626143248_0002_V.zip",
   "name": "DJI_20240626143248_0002_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep2/DJI_20240626143248_0002_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "508849044",
   "sha256": "1008df9164fd152b35e5f6974fe85d33b76b35796dcf9f6a39f2fe0fed7ec313"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep2__DJI_20240626143443_0003_V.zip",
   "name": "DJI_20240626143443_0003_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep2/DJI_20240626143443_0003_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "757876217",
   "sha256": "63298fbb3a06dd9bd7c0d24fbf72cf76295003a22d58b61123ee136ebe8c3535"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep2__DJI_20240626143832_0004_V.zip",
   "name": "DJI_20240626143832_0004_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep2/DJI_20240626143832_0004_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "61702856",
   "sha256": "d3a08d599708096277d5a11289a3cb4d9e477279f21557da35c650641d3d1ffc"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep3__DJI_20260227082000_0001_V.zip",
   "name": "DJI_20260227082000_0001_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep3/DJI_20260227082000_0001_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "347237855",
   "sha256": "b73c1e8389d8c2dfe46bc252315be5a7dce484c80ee303471c7183103fa1a599"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep3__DJI_20260227082115_0002_V.zip",
   "name": "DJI_20260227082115_0002_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep3/DJI_20260227082115_0002_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "1145874099",
   "sha256": "1370ee39d4a24353e6165a96e694e746d67ee87899ff27b86e8b9e97bce28752"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep3__DJI_20260227082707_0003_V.zip",
   "name": "DJI_20260227082707_0003_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep3/DJI_20260227082707_0003_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "326486754",
   "sha256": "926d426a5d4a6f8728d4fac3d35b9ae37c432201dc71444dd62b70cf60552b44"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep3__DJI_20260227083814_0001_V.zip",
   "name": "DJI_20260227083814_0001_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep3/DJI_20260227083814_0001_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "506847021",
   "sha256": "f33b84958abd429c24dd047f4532af4de635ef89ca657b38f8fb842b2074a2ac"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep3__DJI_20260227084135_0002_V.zip",
   "name": "DJI_20260227084135_0002_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep3/DJI_20260227084135_0002_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "695240611",
   "sha256": "7f70c61abda05c62c9a9fa5596656ef24bec16bd27dbf0705028a70b226dcf66"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep3__DJI_20260227084724_0003_V.zip",
   "name": "DJI_20260227084724_0003_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep3/DJI_20260227084724_0003_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "334468909",
   "sha256": "6d6614300db0087e88d208e003d6a456e30cea8490d1bcb7c172265099d13436"
  },
  {
   "@type": "cr:FileObject",
   "@id": "elep3__DJI_20260227084907_0004_V.zip",
   "name": "DJI_20260227084907_0004_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/elep3/DJI_20260227084907_0004_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "266759982",
   "sha256": "24c8da6a153c9772e11aa7ab4cc500fc6f8deb98707b8e69d78113f97b7da82c"
  },
  {
   "@type": "cr:FileObject",
   "@id": "gaze1__DJI_20240624151050_0001_V.zip",
   "name": "DJI_20240624151050_0001_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/gaze1/DJI_20240624151050_0001_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "1124527825",
   "sha256": "f392e4b40560cf904135da409a3b9f8badca4e20ed3d7a80d82e20d446eda8de"
  },
  {
   "@type": "cr:FileObject",
   "@id": "gaze1__DJI_20240624151439_0002_V.zip",
   "name": "DJI_20240624151439_0002_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/gaze1/DJI_20240624151439_0002_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "641896528",
   "sha256": "c2119818213c9dc6482d6477987d16175f044c93827634bed7403cc1aca98426"
  },
  {
   "@type": "cr:FileObject",
   "@id": "gaze1__DJI_20240624151828_0003_V.zip",
   "name": "DJI_20240624151828_0003_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/gaze1/DJI_20240624151828_0003_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "677411778",
   "sha256": "ba9151521024eddf43c2b0d140c05a37599734a537284c4456ce39369cd9a20c"
  },
  {
   "@type": "cr:FileObject",
   "@id": "gaze1__DJI_20240624152216_0004_V.zip",
   "name": "DJI_20240624152216_0004_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/gaze1/DJI_20240624152216_0004_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "637925850",
   "sha256": "eb9eb7034a220269e38870fe8782680aeec2c786ea3e4c77e9d3308eab90707a"
  },
  {
   "@type": "cr:FileObject",
   "@id": "gira1__DJI_20240118140338_0007_V.zip",
   "name": "DJI_20240118140338_0007_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/gira1/DJI_20240118140338_0007_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "578144635",
   "sha256": "e1bada49382c56362b605cf4a7c83ec2a2a891fbf194da8e86302ee03149cd8c"
  },
  {
   "@type": "cr:FileObject",
   "@id": "gira1__DJI_20240118140539_0008_V.zip",
   "name": "DJI_20240118140539_0008_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/gira1/DJI_20240118140539_0008_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "184956165",
   "sha256": "109b73d02d7112848b2df484c6774edd9b198fa5834ca1b15bc421b0283dc682"
  },
  {
   "@type": "cr:FileObject",
   "@id": "gira2__DJI_20250128082221_0004_V.zip",
   "name": "DJI_20250128082221_0004_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/gira2/DJI_20250128082221_0004_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "55065107",
   "sha256": "0f9115dbaa9db609f35eeca9b6e2d4eb2cc8572119465b30997abf8341287978"
  },
  {
   "@type": "cr:FileObject",
   "@id": "gira2__DJI_20250128082625_0006_V.zip",
   "name": "DJI_20250128082625_0006_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/gira2/DJI_20250128082625_0006_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "137405666",
   "sha256": "adb92fe1d596bac8365f5579be2365d3c659df6921f38a16c62fd53be44dde0d"
  },
  {
   "@type": "cr:FileObject",
   "@id": "metadata__segments.csv",
   "name": "segments.csv",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/metadata/segments.csv",
   "encodingFormat": "text/csv",
   "contentSize": "22917",
   "sha256": "95bb2c2b97f41ab7c95f8d8e3640c8c34479831adc282f6026828a7643d34fc1"
  },
  {
   "@type": "cr:FileObject",
   "@id": "metadata__tracks.csv",
   "name": "tracks.csv",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/metadata/tracks.csv",
   "encodingFormat": "text/csv",
   "contentSize": "98848",
   "sha256": "7ec317dc6ad180976c9c9abf1e0c6c0d8fca83d61fd21ce01e12af1f7169ad39"
  },
  {
   "@type": "cr:FileObject",
   "@id": "metadata__videos.csv",
   "name": "videos.csv",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/metadata/videos.csv",
   "encodingFormat": "text/csv",
   "contentSize": "4283",
   "sha256": "45a324310ca57f1e9a987bb395c0fb89ce485a77a4419e71c4b7c71cc5872521"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin1__DJI_20250217155025_0002_D.zip",
   "name": "DJI_20250217155025_0002_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin1/DJI_20250217155025_0002_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "381982612",
   "sha256": "80aa63ff669b215e783d2df738c0ffd4361bb12d3d09c8585618fca7acf3b5bd"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin1__DJI_20250218113812_0013_D.zip",
   "name": "DJI_20250218113812_0013_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin1/DJI_20250218113812_0013_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "443916246",
   "sha256": "22d2f39ca07eade725bbff637127b1ecf46d1da1894d7fafd772895e272b6733"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin1__DJI_20250225095936_0004_D.zip",
   "name": "DJI_20250225095936_0004_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin1/DJI_20250225095936_0004_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "314405591",
   "sha256": "5abf33cd00705268e79a2e467d85c489901ebc9ab17b0a9de262ab3b69f40365"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin1__DJI_20250225100239_0005_D.zip",
   "name": "DJI_20250225100239_0005_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin1/DJI_20250225100239_0005_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "196802118",
   "sha256": "7420edd94ae51697b869fa4da85f1bdd751acee36a074fdf4849e15fbe0ad47e"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin1__DJI_20250228114837_0003_D.zip",
   "name": "DJI_20250228114837_0003_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin1/DJI_20250228114837_0003_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "323219230",
   "sha256": "4c7d33286ce03c646df3636284c43cfe555750c8593784bc64dc0ad6f9464753"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin1__DJI_20250303174548_0001_D.zip",
   "name": "DJI_20250303174548_0001_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin1/DJI_20250303174548_0001_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "294354897",
   "sha256": "0752f2c360eaea20925b5acc8253d025eac473f9a7abe15c0185d663418637f9"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin1__DJI_20250304161638_0001_D.zip",
   "name": "DJI_20250304161638_0001_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin1/DJI_20250304161638_0001_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "309075046",
   "sha256": "a635a8c0f144712b9a815509dd97e67f1b81094e2e4250b46aef40592c08a892"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin1__DJI_20250304161940_0002_D.zip",
   "name": "DJI_20250304161940_0002_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin1/DJI_20250304161940_0002_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "451801123",
   "sha256": "a300334cf144febf620ecda554037635209cad51b3830a5bab68711eacaa03c1"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin1__DJI_20250304162356_0004_D.zip",
   "name": "DJI_20250304162356_0004_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin1/DJI_20250304162356_0004_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "48292507",
   "sha256": "e32d7ccaf31da0f6a31ee4b6a0a93177b7abcc09a38bedb6eb97ddefe5f3687c"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin1__DJI_20250305152334_0023_D.zip",
   "name": "DJI_20250305152334_0023_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin1/DJI_20250305152334_0023_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "161932537",
   "sha256": "b7fd8a02afed1344ee3dd17747487acb2b34e0f4ca43dfeac9d09aea8644ff2e"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin1__DJI_20250305152636_0024_D.zip",
   "name": "DJI_20250305152636_0024_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin1/DJI_20250305152636_0024_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "73647001",
   "sha256": "adea1d7a5ff539b8b9b000ae137de38ac54d107a8a46426e8349b3185ff6e7c8"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin1__DJI_20250312171212_0026_D.zip",
   "name": "DJI_20250312171212_0026_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin1/DJI_20250312171212_0026_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "261443814",
   "sha256": "18e17ec8a4c5d0cdc9164e8195657207890c98462725da98324fb5601635dd9b"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin2__DJI_20250218123544_0017_D.zip",
   "name": "DJI_20250218123544_0017_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin2/DJI_20250218123544_0017_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "515181360",
   "sha256": "144e2d3b55b46062a2795e4e1ed72b55104e63fa85af5ed51f7e60c7acffb6ee"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin2__DJI_20250218175311_0031_D.zip",
   "name": "DJI_20250218175311_0031_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin2/DJI_20250218175311_0031_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "783649356",
   "sha256": "f22fac133daa2975fe98f3753fc7d8696c7e5fa7dc1847ef0d334be0880a2fc1"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin2__DJI_20250218175613_0032_D.zip",
   "name": "DJI_20250218175613_0032_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin2/DJI_20250218175613_0032_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "423245465",
   "sha256": "69ee5334a754c1d0ba373a8d9b602dba310546c8ba30eea702f18286439b593f"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin2__DJI_20250218175916_0033_D.zip",
   "name": "DJI_20250218175916_0033_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin2/DJI_20250218175916_0033_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "181236612",
   "sha256": "528dce5dfe38b6fc72e92d03a23d8c7ca37d4b6284697d7301ead69a789f0391"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin2__DJI_20250218180030_0034_D.zip",
   "name": "DJI_20250218180030_0034_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin2/DJI_20250218180030_0034_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "691107682",
   "sha256": "8135d864ada33fd387b2bf29cbca20d88058ecf6a79ec5eb50f0e8bf4b72259d"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin2__DJI_20250224155903_0001_D.zip",
   "name": "DJI_20250224155903_0001_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin2/DJI_20250224155903_0001_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "567279813",
   "sha256": "35b8b9770395755fda364e86e861365d78af58ede360129cd5c4298ed5720755"
  },
  {
   "@type": "cr:FileObject",
   "@id": "rhin2__DJI_20250225165243_0018_D.zip",
   "name": "DJI_20250225165243_0018_D.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/rhin2/DJI_20250225165243_0018_D.zip",
   "encodingFormat": "application/zip",
   "contentSize": "814136029",
   "sha256": "43570f254cd4123784ff29903488b889164494a7e6ed035f584c381366dcd723"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr1__DJI_20230607092100_0001_V.zip",
   "name": "DJI_20230607092100_0001_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr1/DJI_20230607092100_0001_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "1472566932",
   "sha256": "79f223b1a01a70e7b910573576621da45cc43bb89bd07ae2300c88c592e7dfcf"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr1__DJI_20230607092449_0002_V.zip",
   "name": "DJI_20230607092449_0002_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr1/DJI_20230607092449_0002_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "341396761",
   "sha256": "067219b8c71b03d21ca72642e7d5147e90a6d4e213db586cefa7c46ed6cef7f8"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr1__DJI_20230607092838_0003_V.zip",
   "name": "DJI_20230607092838_0003_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr1/DJI_20230607092838_0003_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "896196064",
   "sha256": "9d3ba3258a39195c031d4c063447201618605428510b278bdb7371ae540a28bb"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr2__DJI_0143.zip",
   "name": "DJI_0143.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr2/DJI_0143.zip",
   "encodingFormat": "application/zip",
   "contentSize": "516537057",
   "sha256": "16db1fed9104e6596f418c0a9f1d334547342356d282575c4c60e461c0d4af8b"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr2__DJI_0144.zip",
   "name": "DJI_0144.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr2/DJI_0144.zip",
   "encodingFormat": "application/zip",
   "contentSize": "652245052",
   "sha256": "6cb3875726b0b310568dd2d8a2d61226f4adc9957a09a80481954ebd4dc2a3ef"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr2__DJI_0145.zip",
   "name": "DJI_0145.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr2/DJI_0145.zip",
   "encodingFormat": "application/zip",
   "contentSize": "1132478313",
   "sha256": "757121973b115cbf61c1d1f739fc335f7b57787b10b52719421528c9868b5ba1"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr2__DJI_0146.zip",
   "name": "DJI_0146.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr2/DJI_0146.zip",
   "encodingFormat": "application/zip",
   "contentSize": "1066421495",
   "sha256": "37d52c241b7bbecb536435523bb916849c48640cc7cb3e1310ffc0a421a2ad92"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr2__DJI_0147_trimmed.zip",
   "name": "DJI_0147_trimmed.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr2/DJI_0147_trimmed.zip",
   "encodingFormat": "application/zip",
   "contentSize": "780637676",
   "sha256": "438b14bf62f0a6fe90545ca287e2b00f4e93e5f7d0bd643dca7e9546252b7839"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr3__DJI_20250802084740_0006_V.zip",
   "name": "DJI_20250802084740_0006_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr3/DJI_20250802084740_0006_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "162344916",
   "sha256": "6026959a2daf072ae7eb82f277cff6a7365f8f7a62cff057c218714d60f96cb7"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr3__DJI_20250802085130_0007_V.zip",
   "name": "DJI_20250802085130_0007_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr3/DJI_20250802085130_0007_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "575029598",
   "sha256": "3815755e02c2cebf985ae2a75e1d8d0bea2dc9e8d4f2459ddf44837385043581"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr3__DJI_20250802085520_0008_V.zip",
   "name": "DJI_20250802085520_0008_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr3/DJI_20250802085520_0008_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "430427677",
   "sha256": "2c922227bc77c964287dd08a221dd366df7035133e04511f0676b96df0d3886e"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr3__DJI_20250802120510_0002_V.zip",
   "name": "DJI_20250802120510_0002_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr3/DJI_20250802120510_0002_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "559554822",
   "sha256": "0dd1e4c8c686406ab9b6fc2dac0c04ac9f2820ff8dfdec20c59e17613533419e"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr3__DJI_20250802120902_0003_V.zip",
   "name": "DJI_20250802120902_0003_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr3/DJI_20250802120902_0003_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "377566510",
   "sha256": "e901aea0a89f3e3de90d75827dc4058923013d7ad6a80579a4c622456b1f2947"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr3__DJI_20250802121253_0004_V.zip",
   "name": "DJI_20250802121253_0004_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr3/DJI_20250802121253_0004_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "636931027",
   "sha256": "a82b94b59382ea47ccf3a5c23f46dd72f6bae4cec4e49397bb76c9fb4885fe7c"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr3__DJI_20250802121643_0005_V.zip",
   "name": "DJI_20250802121643_0005_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr3/DJI_20250802121643_0005_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "1017967810",
   "sha256": "9b74e622ec279a7666559ac0779affa950b5e68d1313fdf0005b9edb28dcea45"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr3__DJI_20250802122032_0006_V.zip",
   "name": "DJI_20250802122032_0006_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr3/DJI_20250802122032_0006_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "725998224",
   "sha256": "0928cdb199acdbfd994b2574a744d2c1f9ed9476f282cc5cee08950040a1cd24"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr3__DJI_20250802125123_0005_V.zip",
   "name": "DJI_20250802125123_0005_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr3/DJI_20250802125123_0005_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "1251609490",
   "sha256": "f44f522437b352e049548e7e16d0e308f12eab01b1b0963e52f2109447d3c178"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr3__DJI_20250802125512_0006_V.zip",
   "name": "DJI_20250802125512_0006_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr3/DJI_20250802125512_0006_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "656740448",
   "sha256": "bfc9b4b9c1470386f3b39d465786979028b10dfaf6b3c5d6fdf99711e381b81f"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr3__DJI_20250802130602_0001_V.zip",
   "name": "DJI_20250802130602_0001_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr3/DJI_20250802130602_0001_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "584861762",
   "sha256": "9ea5d88ece66cabcdd33b1b6e3efb1599e1fad0d3f2f4b642ced660e5becca22"
  },
  {
   "@type": "cr:FileObject",
   "@id": "zebr3__DJI_20250802130951_0002_V.zip",
   "name": "DJI_20250802130951_0002_V.zip",
   "contentUrl": "https://huggingface.co/datasets/3DOMFBK/WildBox/resolve/main/zebr3/DJI_20250802130951_0002_V.zip",
   "encodingFormat": "application/zip",
   "contentSize": "649959383",
   "sha256": "42d413b4fa8ad7752dbd9f764878ede94e53c4c5f70e31473050b36bf330d9a7"
  }
 ],
 "recordSet": [
  {
   "@type": "cr:RecordSet",
   "@id": "images",
   "name": "images",
   "description": "Per-frame image records sourced from the train Omni3D JSON.",
   "field": [
    {
     "@type": "cr:Field",
     "@id": "images/id",
     "name": "id",
     "description": "Unique image identifier.",
     "dataType": "sc:Integer",
     "source": {
      "fileObject": {
       "@id": "WildBox_train.json"
      },
      "extract": {
       "jsonPath": "$.images[*].id"
      }
     }
    },
    {
     "@type": "cr:Field",
     "@id": "images/file_path",
     "name": "file_path",
     "description": "Relative path to the JPEG inside the per-video zip.",
     "dataType": "sc:Text",
     "source": {
      "fileObject": {
       "@id": "WildBox_train.json"
      },
      "extract": {
       "jsonPath": "$.images[*].file_path"
      }
     }
    },
    {
     "@type": "cr:Field",
     "@id": "images/height",
     "name": "height",
     "description": "Image height in pixels (1080 for all WildBox frames).",
     "dataType": "sc:Integer",
     "source": {
      "fileObject": {
       "@id": "WildBox_train.json"
      },
      "extract": {
       "jsonPath": "$.images[*].height"
      }
     }
    },
    {
     "@type": "cr:Field",
     "@id": "images/width",
     "name": "width",
     "description": "Image width in pixels (1920 for all WildBox frames).",
     "dataType": "sc:Integer",
     "source": {
      "fileObject": {
       "@id": "WildBox_train.json"
      },
      "extract": {
       "jsonPath": "$.images[*].width"
      }
     }
    }
   ]
  }
 ],
 "rai:dataLimitations": "WildBox is a relative-scale dataset: per-segment 3D coordinates are normalised by the median camera-space depth of each segment's reconstruction. \nAbsolute metric scale is not recoverable across segments.\n\nGeographic and taxonomic coverage is restricted to seven African savanna species (reticulated giraffe, plains zebra, Grévy's zebra, African bush elephant, Thomson's gazelle, black rhinoceros, white rhinoceros) recorded at three sites: Ol Pejeta Conservancy and Mpala Research Centre in Kenya, and Bristol Zoological Society in the UK.\nGeneralisation to other taxa, biomes, or aerial perspectives outside this distribution has not been validated.\n\nClass distribution is naturally imbalanced and not rebalanced. \nThree classes (giraffe, Grévy's zebra, gazelle) have only 3–4 training videos each; the giraffe validation split contains a single video (110 instances) so per-class metrics for giraffe carry high variance. \nEach frame contains a single species, so within-frame multi-class classification is not exercised.\n\nAnnotations are tracked frame-by-frame within a video segment but identities are not maintained across segments. \nCamera intrinsics are inherited from the source footage and are not separately calibrated.\n\nThis dataset is not recommended for: absolute-scale 3D reconstruction without additional metric supervision; ecological estimation tasks (e.g. abundance estimation) that assume the underlying class frequencies represent true field densities; surveillance, tracking, or behaviour analysis of human subjects; any application that would expose the precise location of endangered individuals.",
 "rai:dataBiases": "WildBox inherits the biases of its data sources and field campaigns.\n\nSite selection bias: data was collected at three locations in Kenya and the UK during 2023–2026. Animals at these sites are habituated to drone presence to varying degrees, and behaviour or appearance under naïve-population conditions may differ.\n\nViewpoint bias: all footage is captured at oblique aerial angles typical of DJI consumer/prosumer drones. Top-down nadir, ground-level, and underwater perspectives are absent.\n\nSpecies and instance imbalance: instance count is decoupled from video-level diversity. Gazelle has 43,280 training instances across just 3 training videos, while rhino has 28,381 instances across 15.\nPer-class results show that video-level diversity is more predictive of detection performance than raw instance count, so models trained without rebalancing may overfit to the recording conditions of\nhigh-instance, low-video-count classes.\n\nTaxonomic granularity is non-uniform: plains zebra and Grévy's zebra are evaluated as separate classes (following KABR convention), while black and white rhinoceros are merged into a single \"rhino\" benchmark class (with a binary species attribute retained in the annotations).\nModels trained on the benchmark labels will not learn to distinguish black from white rhino.\n\nLighting and weather coverage reflects daytime drone operation windows; nighttime, dawn/dusk, and heavy-precipitation conditions are underrepresented.\n\nThe annotation pipeline is model-in-the-loop: WildLIFT proposes 3D bounding boxes and human reviewers refine and confirm them.",
 "rai:personalSensitiveInformation": "WildBox contains only wildlife imagery captured from aerial drones.\nThe dataset includes no human subjects, no personally identifiable information, no health or medical data, and no political or religious content. \nWhere humans are incidentally visible in raw footage, affected frames are excluded during the manual segmentation audit step described in §3 of the paper. \nCars are allowed to be seen in the data.\n\nThe dataset does include endangered species (reticulated giraffe, Grévy's zebra, African bush elephant, both rhino species). \nHowever, no telemetry files are being released.",
 "rai:dataUseCases": "WildBox is intended to measure the geometric structure (position, extent, orientation, depth ordering) of wildlife instances in oblique aerial RGB drone video, expressed as tracked 3D bounding boxes in a\nper-segment scale-normalised camera frame.\n\nValidated use cases (in this paper):\n- Open-vocabulary monocular 3D detection benchmark for wildlife in drone imagery (zero-shot, ground-truth-2D-prompt, and supervised fine-tuning protocols evaluated for OVMono3D-LIFT and DetAny3D).\n- Diagnostic decomposition of monocular 3D error into image-plane, depth, dimension, and pose components (normalised Hausdorff distance).\n- Curriculum-initialised fine-tuning of monocular 3D detectors across taxonomically related subclasses (zebras).\n\nUse cases supported by the data but not validated in this paper:\n- Multi-frame / video-temporal 3D detection (instance IDs are released but not benchmarked).\n- 3D tracking and re-identification (instance IDs are segment-local).\n\nValidation does NOT extend to:\n- Absolute metric 3D reconstruction across segments or flights as the release is per-segment relative-scale.\n- 3D detection of taxa outside the six benchmark classes.\n- 3D detection from non-aerial or non-RGB modalities.\nNote:  WildBox is a method-development benchmark, not a deployment-ready monitoring system.",
 "rai:dataSocialImpact": "Positive impacts: WildBox supports research on monocular 3D perception in a low-cost, deployment-realistic regime (consumer drones, no LiDAR or multi-camera rigs). \nImprovements on this benchmark could feed into non-invasive wildlife monitoring, behavioural ecology, and conservation-relevant applications through wildlife-aware drone autonomy parallel to the research in autonomous driving but for conservation applications. \nThe dataset also extends prior aerial wildlife datasets (which are almost exclusively 2D) into 3D, opening a new line of work on aerial monocular depth estimation.\n\nNegative-impact pathways and mitigations:\n\n- Poaching risk. The dataset includes endangered and critically endangered species (Grévy's zebra, reticulated giraffe, African bush elephant, black and white rhinoceros). To prevent the release from facilitating targeting of individual animals or sites, no GPS coordinates, drone telemetry, or precise location metadata are released with the dataset. \nSite identification is restricted to the conservancy / institution level (Ol Pejeta, Mpala, Bristol Zoological Society) without finer-grained spatial information.\n\n- Mis-deployment in operational settings. Performance on WildBox is reported on a held-out validation split from the same site distribution; cross-site, cross-season, and cross-population generalisation is not measured. The dataset is intended for benchmarking and method development, not as a drop-in component of a deployed conservation pipeline.\n\n- Disturbance to wildlife during data collection. All field campaigns were conducted under valid local permits with documented effort to minimise disturbance (low-altitude flight protocols at the source sites).\n\nAccess terms: the dataset is released under CC-BY 4.0 upon acceptance,\nwith attribution required.",
 "rai:hasSyntheticData": false,
 "prov:wasDerivedFrom": [
  {
   "@id": "https://huggingface.co/datasets/imageomics/KABR-raw-videos",
   "prov:label": "KABR Raw Videos",
   "sc:license": "cc0-1.0"
  }
 ],
 "prov:wasGeneratedBy": [
  {
   "@type": "prov:Activity",
   "prov:type": {
    "@id": "https://www.wikidata.org/wiki/Q4929239"
   },
   "prov:label": "Drone flights and Data collection",
   "sc:description": "- Period: 2023–2026.\n- Geographic scope: Ol Pejeta Conservancy (Laikipia, Kenya), Mpala\n  Research Centre (Laikipia, Kenya — via the public KABR raw video\n  corpus), and Bristol Zoological Society (Bristol, UK).\n- Instruments: DJI consumer/prosumer drones (multiple models across\n  campaigns; oblique RGB at standard drone-monitoring altitudes).\n- Permits: applicable local field permits for each site; data\n  collection followed the host institutions' wildlife-monitoring\n  protocols."
  },
  {
   "@type": "prov:Activity",
   "prov:type": {
    "@id": "https://www.wikidata.org/wiki/Q5227332"
   },
   "prov:label": "Pre-processing",
   "sc:description": "- Each raw video is manually audited and segmented into clips (\"segments\") of up to 200 frames each, with boundaries chosen to keep wildlife presence, viewpoint, and zoom level stable within a clip.\n- Segments containing redundant content, calibration footage, or incidental human appearance are excluded.\n- Frames are sampled at 10 Hz.\n- Per-segment scale normalisation: 3D coordinates within each segment are divided by the median absolute camera-space depth of the segment's reconstruction, preserving box shape, orientation, 3D IoU, and 2D projection within the segment."
  },
  {
   "@type": "prov:Activity",
   "prov:type": {
    "@id": "https://www.wikidata.org/wiki/Q109719325"
   },
   "prov:label": "Data Annotation and Review",
   "sc:description": "- Schema: tracked 3D oriented bounding boxes in the camera frame (X right, Y down, Z forward), each parameterised by position, extent, and a full 3×3 rotation matrix, and tagged with an instance ID consistent within each segment. The benchmark uses six classes (reticulated giraffe, plains zebra, Grévy's zebra, African bush elephant, Thomson's gazelle, rhino — black and white rhino merged with a binary species attribute retained in the annotations).\n- Method: model-in-the-loop. The WildLIFT pipeline (Shukla et al.,2026) produces (i) a dense per-segment 3D pointmap with identity-consistent 3D tracks (WildLIFT-RT) and (ii) oriented 3D bounding-box proposals fitted to each tracked individual (WildLIFT-A). Human reviewers refine and confirm every proposed box frame-by-frame in the WildLIFT-A interface; only human-confirmed boxes enter the released dataset.\n- Annotators: the authors of the paper. No crowdsourcing was used.\n- Quality control: every released annotation has been frame-level reviewed by at least one author."
  }
 ]
}