KnowDemo: Knowledge-Guided Robot Demonstration Generation from Human Videos
A knowledge extraction and reasoning module based on a vision-language model that associates object and action descriptions with inferred task conditions, demonstration references, and permissible execution variations is developed, which translates into executable demonstrations that exhibit multimodal behavior through...
Zhi-Yuan Gao, Yan Zhan, Mohammad Khoshnazar et al.
· 0 citations