Skip to main content
Search

# Contact-Guided Gaze Supervision

1 related articles

Subscribe to updates
  1. arXiv

    Tactile post-training stops reading only instantaneous pressure: CoRA-VTLA recalls recent contact responses and lifts MuJoCo mean success from 50% to 73.33%

    Pretrained vision-language-action models need post-contact correction in precision assembly, yet a single tactile reading cannot tell a tool pressing against an obstruction from contact that has settled. CoRA-VTLA adds a queryable action-response memory on the action-expert side, retrieving recent load changes conditioned on current contact and robot state, and uses contact-guided gaze supervision to pull a tactile semantic token's visual attention toward fingertip contact regions. Across three MuJoCo tasks it raises mean success from 50% for direct tactile fusion with the same sensor inputs to 73.33%, reaches 67% on the UniVTAC optical-tactile tube task, and averages 66.7% across three real-robot tasks.

You’re all caught up