| --- |
| license: cc-by-nc-4.0 |
| task_categories: |
| - video-classification |
| tags: |
| - event-boundary-detection |
| - online-video |
| - kinetics-400 |
| - causal-inference |
| size_categories: |
| - 10K<n<100K |
| pretty_name: Audited Kinetics-GEBD Causal Metadata |
| --- |
| |
| # Audited Kinetics-GEBD Causal Metadata |
|
|
| This metadata-only release converts the publicly released Kinetics-GEBD |
| annotations into an auditable 24 FPS causal training representation. It does |
| not redistribute Kinetics or YouTube video bytes. |
|
|
| ## Splits |
|
|
| | Hub split | Official source file | Records | Meaning | |
| |---|---|---:|---| |
| | `train` | `k400_train_raw_annotation.pkl` | 18,808 | Public GEBD training annotations | |
| | `validation` | `k400_val_raw_annotation.pkl` | 18,815 | Public Kinetics-GEBD validation annotations | |
|
|
| The ICCV 2021 benchmark describes a separate challenge `test` subset sampled |
| from Kinetics-400 train. Its labels are withheld and are not included here. |
| The ICCV 2025 On-GEBD paper explicitly evaluates Kinetics-GEBD on the public |
| validation split because test annotations are unavailable. |
|
|
| ## Annotation processing |
|
|
| The release reproduces the official |
| [`prepare_k400_release.ipynb`](https://github.com/StanLei52/GEBD/blob/5f7e722e0384f9877c75d116e1db72400d2bc58f/data/export/prepare_k400_release.ipynb) |
| behavior, including its input-order and list-mutation quirks: |
|
|
| - discard boundaries in the first or last 0.3 seconds; |
| - merge overlapping gradual-shot ranges; |
| - suppress timestamp boundaries around gradual-shot ranges; |
| - suppress nearby same-type timestamps and prefer shots over events; |
| - convert cleaned timestamps to source-frame indices with |
| `floor(timestamp * source_fps)`, capped at `num_frames - 1`. |
|
|
| The implementation was differentially checked against the official notebook |
| over all 185,518 released rater annotations with zero timestamp-sequence |
| differences. This is intentionally the code behavior: some effective |
| suppression distances are 0.6 seconds even though the paper summarizes a |
| 0.1-second merge rule. |
|
|
| Training uses the first maximum of `f1_consis`, matching `numpy.argmax` in the |
| official loader. Evaluation must compare predictions independently with every |
| retained rater and select the best rater-specific F1. Records with |
| `f1_consis_avg < 0.3` remain present but have `evaluation_eligible=false`. |
|
|
| ## Causal 24 FPS labels |
|
|
| The benchmark does not publish a canonical 24 FPS binary-label file. This |
| release derives one for causal training: |
|
|
| - `sample_count = ceil(duration_seconds * 24)`; |
| - sample `i` represents timestamp `i / 24`; |
| - a sample is positive when it falls inclusively within 0.15 seconds of a |
| cleaned boundary from the highest-consistency training rater. |
|
|
| `boundary_source_frame_indices` and |
| `training_boundary_source_frame_indices` preserve official source-FPS frame |
| coordinates. `training_boundary_sample_indices` and `binary_labels` are the |
| derived 24 FPS projection. These fields must not be conflated. |
|
|
| ## Evaluation |
|
|
| The official metric is relative temporal distance, not the 0.15-second |
| training window. The paper reports F1 at thresholds 0.05 through 0.50; the |
| challenge ranking metric is F1@0.05. Matching is one-to-one and performed |
| separately for each rater, retaining the rater yielding the best F1. |
|
|
| ## Videos |
|
|
| The local acquisition audit resolved and `ffprobe`-validated 37,608 of 37,623 |
| referenced clips: |
|
|
| | Source split | Available | Missing | |
| |---|---:|---:| |
| | train | 18,794 | 14 | |
| | val | 18,814 | 1 | |
|
|
| The remaining 15 references were absent from all inspected public mirrors, |
| the official CVDF Kinetics-400 archives, and accessible original-video paths. |
| See [`VIDEO_ACQUISITION.md`](VIDEO_ACQUISITION.md) and `manifests/` for exact |
| availability and missing-reference records. |
|
|
| ```bash |
| uv run scripts/build_official_on_gebd_dataset.py |
| uv run scripts/acquire_official_gebd_videos.py --workers 4 |
| ``` |
|
|
| ## Rights |
|
|
| The GEBD annotation release is CC BY-NC 4.0 and the official repository code |
| is MIT-licensed. Neither license grants redistribution rights for the |
| underlying Kinetics/YouTube videos. A gated or private Hub repository does not |
| change those rights, so this release contains metadata and manifests only. |
|
|
| ## Sources |
|
|
| - [Generic Event Boundary Detection, ICCV 2021](https://openaccess.thecvf.com/content/ICCV2021/papers/Shou_Generic_Event_Boundary_Detection_A_Benchmark_for_Event_Segmentation_ICCV_2021_paper.pdf) |
| - [Online Generic Event Boundary Detection, ICCV 2025](https://openaccess.thecvf.com/content/ICCV2025/papers/Jung_Online_Generic_Event_Boundary_Detection_ICCV_2025_paper.pdf) |
| - [Official GEBD repository](https://github.com/StanLei52/GEBD/tree/5f7e722e0384f9877c75d116e1db72400d2bc58f) |
| - [Official challenge evaluator](https://github.com/StanLei52/GEBD/blob/5f7e722e0384f9877c75d116e1db72400d2bc58f/Challenge_eval_Code/eval.py) |
| - [Official Kinetics downloads](https://github.com/cvdfoundation/kinetics-dataset) |
|
|
