robotics
Robot Motion Becomes a Short Vocabulary
KiRoSe compresses kinematic trajectories into sparse motion tokens that an open language model can segment without ingesting every camera stream.
Summary
KiRoSe compresses kinematic trajectories into sparse motion tokens that an open language model can segment without ingesting every camera stream.
Robot demonstrations are usually segmented by hand or by feeding expensive multi-camera video into multimodal models. KiRoSe instead learns a discrete vocabulary of localized kinematic events from joint and motion trajectories, then trains an open language model to interpret those token sequences. Against human labels, the authors report higher F1 than baseline segmentation methods, zero-shot transfer to unseen robot bodies and egocentric human motion, plus progress tracking and reversal detection on an unseen robot. The abstract does not provide a universal margin, and the vocabulary learns from available subtask annotations. It is a compact companion to video, not evidence that motion alone captures every semantic boundary.
Why it matters
KiRoSe compresses kinematic trajectories into sparse motion tokens that an open language model can segment without ingesting every camera stream.
Limits and context
- The abstract does not provide a universal margin, and the vocabulary learns from available subtask annotations.
- It is a compact companion to video, not evidence that motion alone captures every semantic boundary.
Key claims
KiRoSe compresses kinematic trajectories into sparse motion tokens that an open language model can segment without ingesting every camera stream.
Qualification: The abstract does not provide a universal margin, and the vocabulary learns from available subtask annotations.
Evidence: source-2026-10-26-014
Sources
- arXiv preprint 2610.31801, original v1arXiv · primary research
Corrections
No corrections have been recorded for this story.