| --- |
| license: other |
| pretty_name: RETRACK3D |
| tags: |
| - point-tracking |
| - 3d |
| - egocentric |
| - depth |
| - pytorch |
| --- |
| |
| # RETRACK3D |
|
|
| Test trial for dataset export and dataloader validation. Not an official release. |
|
|
| ## Storage and continuity |
|
|
| ```text |
| loader/ standalone Python reader |
| examples/read_pair.py read 96 + 96 as a 192-frame sample |
| docs/FORMAT.md schema, identities, coordinates and sampling |
| validation/ parity, data checks and training reports |
| integration/ InfD4RT adapter and resumable training entry point |
| xperience/chunks96/v2/ test data payload |
| manifest.json |
| manifest.sha256 |
| chunks/<capture>_ep<n>/ |
| 0000000-0000096-<hash>.h5 |
| 0000096-0000192-<hash>.h5 |
| ... |
| ``` |
|
|
| Frame intervals are half-open. Adjacent files retain episode-global frame IDs, |
| object-track IDs and the same original world coordinate system. The reader |
| concatenates raw arrays before sampling queries and estimating gravity once for |
| the full window. Tail files contain 96–191 frames to avoid padding or dropping |
| a short tail. Whole episodes shorter than 96 frames are preserved too; configure |
| a shorter sample window to use episodes shorter than the default 48 frames. Images use H.264 CRF 18; depth, geometry, masks and IMU use lossless |
| compression at their source/export precision. |
|
|
| Each supervision head has an independent validity mask. Unknown visibility is |
| not an invisible label. Static background targets through occlusion follow a |
| rigidity assumption; these are pseudo-labels, not independent hidden-point GT. |
|
|
| ## Reader |
|
|
| Install `loader/requirements.txt`; add `loader/` to `PYTHONPATH`, or run |
| `examples/read_pair.py`, which finds the bundled loader automatically. The reader |
| accepts local paths, HTTP and fsspec URIs, including Hugging Face. It fetches only |
| the manifest at initialization and shares a checksum-verified, bounded disk cache |
| between workers on each node. Warm cache hits download zero bytes. |
|
|
| With an authorized local payload: |
|
|
| ```bash |
| python examples/read_pair.py --manifest /path/to/manifest.json \ |
| --episode 0073a4be-ec41-47f3-92d6-f63cf89aec9b_ep8.h5 --start 960 |
| ``` |
|
|
| See [FORMAT.md](docs/FORMAT.md) for all fields and the InfD4RT training interface. |
| [Training checks](validation/training.json) exercised the 1.535B-parameter |
| model at 48 frames / 224 pixels on an A40, including checkpoint resume. |
| A separate small model trained on joined 192-frame inputs. All tested steps had |
| finite losses/gradients and actual parameter updates; |
| they do not establish convergence or annotation accuracy. |
|
|
| ## Source and terms |
|
|
| Source: [Xperience-10M](https://huggingface.co/datasets/ropedia-ai/xperience-10m), |
| Ropedia. Its data card specifies controlled access and an external agreement. |
| Underlying source-data rights and terms continue to apply; this repository does |
| not relicense those recordings under MIT. The payload is distributed under manual gated access. |
|
|
| ## Test export |
|
|
| The current test export is under `xperience/chunks96/v2/`. |
| The `v2-preview` directory contains the earlier test subset. |
| Windows and queries are sampled at runtime, including across adjacent files. |
|
|
| ## Remote test reader |
|
|
| Use `hf://datasets/2inf/RETRACK3D/xperience/chunks96/v2/manifest.json` with an approved |
| Hugging Face account. Pin the repository revision for reproducible training. |
| The default query allocation is background 50%, object-track 35%, hand 15%; |
| missing pools redistribute the quota without weakening validity masks. |
| Object-track type is not a guarantee of independent motion or a manipulated object. |
| See `validation/full_upload.json` for upload integrity checks. |
|
|
