Update README for flattened single-split structure with embedded videos and inline boxes

README.md

@@ -16,85 +16,93 @@ configs:
 - config_name: default
   data_files:
   - split: train
-    path: data/train-*
-dataset_info:
-  features:
-  - name: video
-    dtype:
-      video:
-        decode: false
-  - name: video_id
-    dtype: string
-  - name: patient_id
-    dtype: int64
-  - name: boxes
-    struct:
-    - name: track_id
-      list: int64
-    - name: frame
-      list: int64
-    - name: label
-      list: string
-    - name: xtl
-      list: float32
-    - name: ytl
-      list: float32
-    - name: xbr
-      list: float32
-    - name: ybr
-      list: float32
-    - name: occluded
-      list: int64
-    - name: outside
-      list: int64
-    - name: keyframe
-      list: int64
-    - name: z_order
-      list: int64
-  - name: n_boxes
-    dtype: int64
-  - name: histological_type
-    dtype: string
-  - name: num_frames
-    dtype: int64
-  - name: light_mode
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 4327196551
-    num_examples: 173
-  download_size: 4327526548
-  dataset_size: 4327196551
+    path: data/train-*.parquet
 ---
 
 # Cystoscopy Tumor Detection
 
-Cystoscopic video dataset with per-frame `tumor` bounding-box annotations, paired with patient-level clinical metadata.
+Cystoscopic video dataset with per-frame `tumor` bounding-box annotations,
+paired with patient-level clinical metadata. Flattened into a single `train`
+split with one row per video; videos are embedded in the Parquet shards as HF
+`Video` features and bounding boxes are stored inline.
 
-## Contents
+## Why no splits
 
-| Path | Description |
-|---|---|
-| `data/videos/` | 173 raw cystoscopy `.mp4` tracks (~4 GB). Filename pattern `P{NN}_cystoscopy_track_{NNN}.mp4`. |
-| `data/annotations.parquet` | Long-format bounding boxes: one row per box (69 108 rows, 444 tracks). Columns: `video_id, patient_id, track_id, frame, label, xtl, ytl, xbr, ybr, occluded, outside, keyframe, z_order`. |
-| `data/clinical_info.parquet` | One row per patient (n=30). Columns: `patient_id, histological_type, num_frames, light_mode`. |
-| `splits/{train,val,test}.txt` | Patient-grouped video-id lists (seeded shuffle, seed=42). 24 / 3 / 3 patients → 132 / 18 / 23 videos. No patient appears in more than one split. |
+This dataset ships **one `train` split only**. Use the `patient_id` column to
+build your own patient-grouped splits to prevent leakage. 30 unique patients.
 
-## Notes
+## Columns
 
-- Box coordinates are absolute pixel coordinates in the source video frames (CVAT format).
-- `annotated_videos/` (boxes burned into the pixels) is **not** shipped — it is fully reproducible from `annotations.parquet` + `data/videos/` via the included `overlay_boxes.ipynb`.
-- Splits are by **patient** to prevent leakage; adjust by editing `splits/*.txt`.
+| Column | Type | Description |
+|---|---|---|
+| `video` | `Video` | Embedded video bytes (`decode=False` — returns `{bytes, path}`; cast to `Video(decode=True)` to decode frames, requires `torchcodec`). |
+| `video_id` | `string` | Filename stem, e.g. `P000_cystoscopy_track_000`. |
+| `patient_id` | `int64` | Patient id — use this for group-aware splitting. |
+| `boxes` | `struct` | Per-box annotations as parallel lists (see below). `n = len(boxes["frame"])` boxes for this video. |
+| `n_boxes` | `int64` | Number of boxes for this video. |
+| `histological_type` | `string` | Patient-level histology (e.g. `Urothelial carcinoma pTaLG`). |
+| `num_frames` | `int64` | Patient-level total annotated frames. |
+| `light_mode` | `string` | Imaging light mode (e.g. `CLARA + CHROMA`). |
+
+### `boxes` struct fields
+
+Each field is a list of length `n_boxes`; index `i` across all fields
+describes one box.
+
+| Field | Type | Description |
+|---|---|---|
+| `track_id` | `int64` | Annotation track within the video. |
+| `frame` | `int64` | Frame number the box belongs to. |
+| `label` | `string` | Box label (always `tumor`). |
+| `xtl`, `ytl`, `xbr`, `ybr` | `float32` | Absolute pixel coordinates (CVAT format). |
+| `occluded` | `int64` | Occlusion flag. |
+| `outside` | `int64` | Outside flag. |
+| `keyframe` | `int64` | Keyframe flag. |
+| `z_order` | `int64` | Z-order. |
+
+## Contents
+
+- 173 cystoscopy `.mp4` videos (~4 GB, embedded).
+- 69 108 bounding boxes across 444 annotation tracks.
+- 30 patients with clinical metadata.
 
 ## Loading
 
 ```python
-import pandas as pd, pathlib
-ann = pd.read_parquet("hf://datasets/milkyroad/B/data/annotations.parquet")
-cli = pd.read_parquet("hf://datasets/milkyroad/B/data/clinical_info.parquet")
-train_ids = pathlib.Path("hf://datasets/milkyroad/B/splits/train.txt").read_text().split()
+from datasets import load_dataset
+ds = load_dataset("milkyroad/B", split="train")
+# video is not decoded by default (no torchcodec required to load)
+print(ds[0]["video"])          # {'bytes': ..., 'path': 'P000_cystoscopy_track_000.mp4'}
+print(ds[0]["n_boxes"])        # 233
+boxes = ds[0]["boxes"]
+print(boxes["frame"][0], boxes["label"][0], boxes["xtl"][0])
+```
+
+### Decode video frames
+
+To decode frames, install `torchcodec` and cast the column:
+
+```python
+from datasets import Video
+ds = ds.cast_column("video", Video(decode=True))
+```
+
+## Group-aware split example
+
+```python
+import random
+ds = load_dataset("milkyroad/B", split="train")
+pids = sorted({r["patient_id"] for r in ds})
+random.Random(42).shuffle(pids)
+n_test, n_val = 3, 3
+test_pids = set(pids[:n_test])
+val_pids = set(pids[n_test:n_test + n_val])
+train = ds.filter(lambda r: r["patient_id"] not in test_pids and r["patient_id"] not in val_pids)
+val   = ds.filter(lambda r: r["patient_id"] in val_pids)
+test  = ds.filter(lambda r: r["patient_id"] in test_pids)
 ```
 
-## Reproducibility
+## Notes
 
-Splits generated with `random.Random(42)` over the sorted patient-id list (0–29).
+- Box coordinates are absolute pixel coordinates in the source video frames (CVAT format).
+- Splits should be by **patient** to prevent leakage; `patient_id` is provided for this purpose.