Ep#97: Cross-Embodiment Robot Manipulation via a Unified Hand Action Space

Podcast: RoboPapers · Chris Paxton and Michael Cho

Opublikowano: 2026-08-14

Representing actions for robot hands is actually very challenging. Robot hands have different sizes and shapes, and different degrees of freedom. Representing their actions purely as joint states — as we naively would — leads to less transfer across the wide variety of available hands. Instead, could we design a universal hand action space to allow robots to learn cross-embodiment dexterous manipulation more easily? Luis Felipe Casas and Yu Xiang come on to tell us about UHAS: the Unified Hand Action Space, a sphere-based action representation which enables cross-ebmodiment learning of dexterous manipulation skills. Watch Episode #97 of Robo...