arXiv AI By Wenda Yu, Tianshi Wang, Fengling Li, Xin Li, Jingjing Li, Lei Zhu

ActFovea: Runtime Safeguarding for VLA Policies via Spatiotemporal Visual-Action Consistency

Read the original on arXiv AI →

arXiv:2607. 29169v1 Announce Type: cross Abstract: Vision-language-action (VLA) policies achieve strong performance in robotic manipulation but remain vulnerable to runtime disturbances that break the temporal alignment among visual observations, robot states, and executed actions.

Summary generated by The Flow from the publisher's feed. The full article lives at arXiv AI.