---
library_name: braindecode
license: cc-by-nc-sa-4.0
tags: [emg, semg, hand-pose, pose-estimation, braindecode, emg2pose]
pipeline_tag: other
---

# vemg2pose — emg2pose Tracking baseline

Meta's released `tracking_vemg2pose.ckpt` from the
[emg2pose benchmark](https://github.com/facebookresearch/emg2pose), rehosted with
[braindecode](https://braindecode.org)'s parameter names:

```python
from braindecode.models import VEMG2Pose

model = VEMG2Pose.from_pretrained("braindecode/VEMG2Pose-emg2pose-tracking")
```

16-channel sEMG at 2 kHz in, 20 joint angles per sample out. The encoder uses **valid**
convolutions and consumes a left context of **1790** samples, so windows must be longer
than that; the paper trains on 11,790 (10,000 + 1790).

## Configuration

`decoder="lstm"`, `parameterization="velocity"`. The recurrent decoder emits one velocity per joint, integrated from a supplied initial pose. Call `model(x, y0)`; this checkpoint was trained with the ground-truth initial pose.

## Provenance

Not retrained — the authors' checkpoint with parameter names rewritten. Loaded into
`braindecode.models.VEMG2Pose` and run against the reference `emg2pose.pose_modules`
implementation on the same input, the outputs are **bit-identical** (max absolute
difference 0.0), with all 68 tensors mapped.

## License

Weights are Meta's, under **CC BY-NC-SA 4.0** (non-commercial, share-alike), carried
over unchanged. The UmeTrack hand model behind the labels is CC BY-NC 4.0.

- Salter et al. (2024), *emg2pose*, NeurIPS Datasets and Benchmarks.
  [arXiv:2412.02725](https://arxiv.org/abs/2412.02725)