HRT1 accepted to RA-L: a robot learns mobile manipulation from a single human video

YuXiang_IRVL · x · 2026-11-18

UT Dallas IRVL's HRT1 paper has been accepted to IEEE RA-L, with presentation at ICRA 2027. Key idea: unlike prior robot-learning work that needs lots of data and retraining, HRT1 lets a robot learn mobile manipulation tasks from just one human demonstration video via trajectory transfer — and execute them in new environments with relocated objects. The team frames it as an efficient use of human video data for robot generalization.

Original post →

More from Embodied

Embodied channel →