WristCompass: Kinematic Coupling as a Learnable Visual Concept for Ego-Camera Orientation
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866914615422615552 |
|---|---|
| author | Nair, Varun Baradwaj, Vidyut He, Jiahang Singh, Anya Relan, Jai Happi, Cabrel |
| author_facet | Nair, Varun Baradwaj, Vidyut He, Jiahang Singh, Anya Relan, Jai Happi, Cabrel |
| contents | Recovering ego-camera orientation from manipulation video is a prerequisite for disentangling hand motion from camera motion, a key step in imitation learning from egocentric demonstrations. The obvious approach, inferring orientation from scene geometry, fails when hands occlude the frame: VGGT, a 1B-parameter scene reconstruction model, scores worse than a constant predictor on the TACO benchmark. We identify an alternative visual concept that is present precisely when scene geometry is absent: kinematic coupling dynamics, the structured physical relationship between wrist motion and camera orientation imposed by the arm-shoulder-head chain. We find that this concept is compact (4D inter-wrist features outperform 126D full hand keypoints), temporal (requiring a GRU over short windows rather than per-frame retrieval), and physically grounded (transferring zero-shot across datasets because it is rooted in anatomy rather than scene appearance). Trained only on tabletop manipulation, WristCompass transfers zero-shot to Epic Kitchens cooking video, achieving 14.3$^\circ$ median geodesic error and approaching the performance of a 1B-parameter scene model at 200K GRU parameters. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_30671 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | WristCompass: Kinematic Coupling as a Learnable Visual Concept for Ego-Camera Orientation Nair, Varun Baradwaj, Vidyut He, Jiahang Singh, Anya Relan, Jai Happi, Cabrel Computer Vision and Pattern Recognition Robotics Recovering ego-camera orientation from manipulation video is a prerequisite for disentangling hand motion from camera motion, a key step in imitation learning from egocentric demonstrations. The obvious approach, inferring orientation from scene geometry, fails when hands occlude the frame: VGGT, a 1B-parameter scene reconstruction model, scores worse than a constant predictor on the TACO benchmark. We identify an alternative visual concept that is present precisely when scene geometry is absent: kinematic coupling dynamics, the structured physical relationship between wrist motion and camera orientation imposed by the arm-shoulder-head chain. We find that this concept is compact (4D inter-wrist features outperform 126D full hand keypoints), temporal (requiring a GRU over short windows rather than per-frame retrieval), and physically grounded (transferring zero-shot across datasets because it is rooted in anatomy rather than scene appearance). Trained only on tabletop manipulation, WristCompass transfers zero-shot to Epic Kitchens cooking video, achieving 14.3$^\circ$ median geodesic error and approaching the performance of a 1B-parameter scene model at 200K GRU parameters. |
| title | WristCompass: Kinematic Coupling as a Learnable Visual Concept for Ego-Camera Orientation |
| topic | Computer Vision and Pattern Recognition Robotics |
| url | https://arxiv.org/abs/2605.30671 |