Rewrite dataset card: YAML metadata, schema table, category counts, loading example

README.md

@@ -1,15 +1,53 @@
+---
+license: other
+license_name: academic-non-commercial
+task_categories:
+- visual-question-answering
+- video-text-to-text
+language:
+- en
+tags:
+- video
+- human-motion
+- 3d-pose
+- multiple-choice
+- benchmark
+size_categories:
+- 100K<n<1M
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: train.jsonl
+  - split: validation
+    path: val.jsonl
+  - split: test_emdb
+    path: test_emdb.jsonl
+  - split: test_rich
+    path: test_rich.jsonl
+  - split: test_egobody
+    path: test_egobody.jsonl
+  - split: test_3dpw
+    path: test_3dpw.jsonl
+  - split: test_sloper4d
+    path: test_sloper4d.jsonl
+---
+
 # HumanMoveVQA
 
-Human motion video question answering. Questions are generated from 3D human pose
-estimated with [PromptHMR](https://github.com/yufu-wang/PromptHMR) and captioned with
-MotionScript. This is the **v6.3** release.
+Multiple-choice question answering about human motion in video. Questions are
+generated from 3D human pose estimated with
+[PromptHMR](https://github.com/yufu-wang/PromptHMR) and captioned with MotionScript.
+
+This is the **v6.3** release. It covers five source datasets: EMDB, RICH, EgoBody,
+3DPW, and SLOPER4D.
 
 ## Files
 
-| File | Rows | Datasets |
-|------|-----:|----------|
-| `train.jsonl` | 86,797 | RICH, EgoBody, EMDB, 3DPW, SLOPER4D |
-| `val.jsonl` | 11,081 | RICH, EgoBody, EMDB, 3DPW, SLOPER4D |
+| File | Rows | Source dataset |
+|------|-----:|----------------|
+| `train.jsonl` | 86,797 | all five |
+| `val.jsonl` | 11,081 | all five |
 | `test_emdb.jsonl` | 768 | EMDB |
 | `test_rich.jsonl` | 2,218 | RICH |
 | `test_egobody.jsonl` | 6,994 | EgoBody |
@@ -19,7 +57,7 @@ MotionScript. This is the **v6.3** release.
 `val` is the held-out in-domain split used for checkpoint selection. The `test_*`
 files are the evaluation benchmark.
 
-The train, val, and test splits share no videos and no questions.
+Train, val, and test share no videos and no questions, in every source dataset.
 
 ## Schema
 
@@ -44,14 +82,34 @@ One row per question:
 }
 ```
 
-- `data` is the frame directory relative to the dataset root. See Video Sources below.
-- `category` is one of `numerical`, `comparative`, `dominant`, `temporal`,
-  `existence`, `ordering`, `trajectory_affordance`.
-- `answer` is `"<letter>. <option text>"`. `D` is `null` for two-option questions.
-- `source_video_start` and `source_video_end` are `M:SS.mmm`. They are read from the
-  video file duration, not from the caption metadata.
-- For EgoBody entries whose `video_id` carries a `_partNN` suffix, `source_video_id`
-  is the full recording and the timestamps give the segment's position within it.
+| Field | Meaning |
+|---|---|
+| `id` | Row identifier, unique within a file |
+| `source_dataset` | One of `EMDB`, `RICH`, `EgoBody`, `3DPW`, `SLOPER4D` |
+| `source_video_id` | Full recording. Differs from `video_id` for EgoBody `_partNN` segments |
+| `source_video_start` | Segment start within the full recording, `M:SS.mmm` |
+| `source_video_end` | Segment end within the full recording, `M:SS.mmm` |
+| `video_id` | The video the question was generated from |
+| `data` | Frame directory relative to the dataset root |
+| `category` | Question type, see below |
+| `question` | Question stem |
+| `A`, `B`, `C`, `D` | Options. `C` and `D` are `null` for two-option questions |
+| `answer` | `"<letter>. <option text>"` |
+
+`source_video_start` and `source_video_end` are read from the video file duration,
+not from the caption metadata.
+
+### Categories
+
+| Category | Count (train + val) | What it asks |
+|---|---:|---|
+| `existence` | 20,891 | Whether a motion event occurred |
+| `numerical` | 16,070 | How many times an event occurred |
+| `comparative` | 14,021 | Which of two directions has greater magnitude |
+| `dominant` | 13,977 | Which direction dominates overall |
+| `temporal` | 13,939 | When an event occurred, or its speed |
+| `trajectory_affordance` | 10,946 | Path shape and what the motion affords |
+| `ordering` | 8,034 | The order in which events occurred |
 
 ## Changes in v6.3
 
@@ -63,151 +121,126 @@ One row per question:
 
 The previous release is available at the `v6.2-legacy` tag.
 
-## Video Sources
+## Source datasets
 
-This section explains how to obtain the source videos for each dataset. The `data`
-field in each entry points to the relative path of the image frames within the
-dataset's directory structure.
+All five are academic, non-commercial use only. Each requires its own request or
+download step. The videos are **not** redistributed here; only the questions and
+answers are.
 
-**Frame rate:** All datasets use **24 fps** when assembling image frames into video,
-except 3DPW and SLOPER4D, which use **30 fps**.
+| Dataset | Videos (train / val / test) | Frame rate | Downsample | Access |
+|---|---|---|---|---|
+| RICH | 701 / 83 / 32 | 24 fps | 8 | [rich.is.tue.mpg.de](https://rich.is.tue.mpg.de/) |
+| EgoBody | 300 / 39 / 54 | 24 fps | 4 | [sanweiliti.github.io/egobody](https://sanweiliti.github.io/egobody/egobody.html) |
+| EMDB | 57 / 7 / 11 | 24 fps | 2 | [emdb.ait.ethz.ch](https://emdb.ait.ethz.ch/) |
+| 3DPW | 40 / 4 / 15 | 30 fps | 1 | [3DPW project page](https://virtualhumans.mpi-inf.mpg.de/3DPW/) |
+| SLOPER4D | 13 / 2 / 15 | 30 fps | 1 | [SLOPER4D project page](https://github.com/climbingdaily/SLOPER4D) |
 
-**Spatial downsampling** (width/height divided by factor when converting frames to video):
+Frame rate is the rate at which image frames are assembled into video, measured from
+the frame count and the video duration. Downsample is the factor applied to the
+source image width and height when rendering the video.
 
-| Dataset | Downsample factor |
-|---------|------------------|
-| EMDB | 2 |
-| RICH | 8 |
-| EgoBody | 4 |
-| 3DPW | 1 |
-| SLOPER4D | 1 |
+Split properties:
 
----
+- EgoBody and SLOPER4D splits are group-disjoint. EgoBody groups by recording, so
+  views of the same recording never straddle a split. SLOPER4D groups by subject, and
+  the benchmark holds three distinct subjects.
+- SLOPER4D clips pass a horizontal alignment gate (XZ RMS <= 2 m against LiDAR GT).
+  Clips failing the gate, and RGB-only clips with no GT, are excluded.
+- 3DPW is 30 fps while the other three original datasets are 24 fps, so 3DPW results
+  are reported separately in the paper.
+
+## Video sources and path mapping
 
-## EMDB
+The `data` field points at the frame directory, not the video file. The layout
+differs per dataset.
 
-**License:** Academic / non-commercial use only. Requires institutional approval.
+### EMDB
 
 **Request access:** [https://emdb.ait.ethz.ch/](https://emdb.ait.ethz.ch/)
-Fill out the form with an institutional email address. Access to the download is granted after approval.
+Fill out the form with an institutional email address. Access is granted after approval.
 
 **Download:** Once approved, download all partition folders `P0`, `P1`, ..., `P9`.
 
-**Data path mapping:**
-The `data` field gives the path to the image frames relative to the EMDB root, e.g.:
-
 ```
 "data": "P0/07_outdoor_push_ups/images/"
+  ->  <EMDB_ROOT>/P0/07_outdoor_push_ups/images/
 ```
 
-corresponds to:
-
-```
-<EMDB_ROOT>/P0/07_outdoor_push_ups/images/
-```
-
-> **Note on partition nesting:** Some archives unzip into a doubled folder, so the
-> frames land at `<EMDB_ROOT>/P2/P2/<sequence>/images/`. Check your local layout and
-> adjust the prefix if needed.
+> **Partition nesting:** Some archives unzip into a doubled folder, so the frames
+> land at `<EMDB_ROOT>/P2/P2/<sequence>/images/`. Check your local layout and adjust
+> the prefix if needed.
 
----
-
-## RICH
-
-**License:** Academic / non-commercial use only.
+### RICH
 
 **Request access:** [https://rich.is.tue.mpg.de/](https://rich.is.tue.mpg.de/)
 
-**Download:** From the download page, download the JPEG image files for all three splits: `train`, `val`, and `test`. The `data` field in each entry indicates which split the sequence belongs to, so only the relevant split files are strictly required for a given JSONL.
-
-**Data path mapping:**
-The `data` field gives `<split>/<scene>/<camera>/`, e.g.:
+**Download:** From the download page, download the JPEG image files for all three
+splits: `train`, `val`, and `test`. The `data` field indicates which split a sequence
+belongs to, so only the relevant split files are strictly required.
 
 ```
 "data": "val/Pavallion_003_yoga1/cam_10/"
+  ->  <RICH_ROOT>/val/Pavallion_003_yoga1/cam_10/
 ```
 
-corresponds to:
-
-```
-<RICH_ROOT>/val/Pavallion_003_yoga1/cam_10/
-```
-
-> **Note on the train prefix:** Some copies of the RICH train split are nested one
-> level deeper, so the frames land at `<RICH_ROOT>/train2/train/<scene>/<camera>/`.
-> The `data` field reflects the layout the frames were read from. Adjust the prefix
-> to match your local copy.
-
----
-
-## EgoBody
-
-**License:** Academic / non-commercial use only.
+> **Train prefix:** Some copies of the RICH train split are nested one level deeper,
+> so the frames land at `<RICH_ROOT>/train2/train/<scene>/<camera>/`. The `data` field
+> reflects the layout the frames were read from. Adjust the prefix to match your copy.
 
-**Request access & Download:** [https://sanweiliti.github.io/egobody/egobody.html](https://sanweiliti.github.io/egobody/egobody.html)
-Sign the license and download the `kinect_color` folder, which contains per-recording image frames organised by camera (`master`, `sub_1`, `sub_2`).
+### EgoBody
 
-**Data path mapping:**
-The `data` field gives `<recording>/<camera>/`, e.g.:
+**Request access & download:** [https://sanweiliti.github.io/egobody/egobody.html](https://sanweiliti.github.io/egobody/egobody.html)
+Sign the license and download the `kinect_color` folder, which holds per-recording
+image frames organised by camera (`master`, `sub_1`, `sub_2`).
 
 ```
 "data": "recording_20211004_S19_S06_01/sub_1/"
+  ->  <EGOBODY_ROOT>/kinect_color/recording_20211004_S19_S06_01/sub_1/
 ```
 
-corresponds to:
-
-```
-<EGOBODY_ROOT>/kinect_color/recording_20211004_S19_S06_01/sub_1/
-```
-
-> **Note on split videos:** Some entries have a `video_id` with a `_partXX` suffix
-> (e.g. `recording_20211004_S19_S06_01_sub1_part02`). This indicates the QA was
+> **Split videos:** Some entries have a `video_id` with a `_partNN` suffix, for
+> example `recording_20211004_S19_S06_01_sub1_part02`. This means the questions were
 > generated on a temporal segment of the full recording. `source_video_id` gives the
 > full recording, and `source_video_start` / `source_video_end` give the time range
 > within it. All image frames are still found under the same `data` path.
 
----
-
-## 3DPW
-
-**License:** See the [3DPW project page](https://virtualhumans.mpi-inf.mpg.de/3DPW/).
+### 3DPW
 
 **Download:** Download the image sequence archive (`imageFiles.zip`) and extract it.
 
-**Data path mapping:**
-The `data` field gives `<sequence>/`, e.g.:
-
 ```
 "data": "courtyard_backpack_00/"
+  ->  <3DPW_ROOT>/imageFiles/courtyard_backpack_00/
 ```
 
-corresponds to:
-
-```
-<3DPW_ROOT>/imageFiles/courtyard_backpack_00/
-```
-
-> **Note on clip videos:** Some entries have a `video_id` with a `_clipN` suffix
-> (e.g. `downtown_bus_00_clip0`). The frames live under the unsuffixed sequence
-> directory, so `data` is `downtown_bus_00/`.
-
----
-
-## SLOPER4D
+> **Clip videos:** Some entries have a `video_id` with a `_clipN` suffix, for example
+> `downtown_bus_00_clip0`. The frames live under the unsuffixed sequence directory,
+> so `data` is `downtown_bus_00/`.
 
-**License:** See the [SLOPER4D project page](https://github.com/climbingdaily/SLOPER4D).
+### SLOPER4D
 
 **Download:** Download the SLOPER4D dataset and extract the RGB clips.
 
-**Data path mapping:**
-The `data` field gives `<sequence>/<clip>/`, e.g.:
-
 ```
 "data": "seq002_football_001/seq002_football_001_c000/"
+  ->  <SLOPER4D_ROOT>/clips/seq002_football_001/seq002_football_001_c000/
 ```
 
-corresponds to:
+## Loading
 
+```python
+from datasets import load_dataset
+
+ds = load_dataset("humanmovevqa/HumanMoveVQA", "default")
+print(ds["test_emdb"][0])
 ```
-<SLOPER4D_ROOT>/clips/seq002_football_001/seq002_football_001_c000/
-```
+
+## Citation
+
+If you use this dataset, please cite the HumanMoveVQA paper and the source datasets
+you use. See the paper for the full reference list.
+
+## Generation pipeline
+
+The code that builds this dataset is at
+[github.com/humanmovevqa/HumanMove_VQA](https://github.com/humanmovevqa/HumanMove_VQA).