arXiv Robotics — research abstracts· Huang Huang, Sriram Yenamandra, Arjun Majumdar, Elie Aljalbout, Tushar Nagarajan, Tsung-Yen Yang, Akshara Rai, Michael Rabbat, Li Fei-Fei, Jiajun Wu, Tingfan Wu, Franziska Meier·· 1 days agoEditorial score63
Cross-Embodiment Robot Foundation World Models with Latent Actions
Cross-Embodiment Robot Foundation World Models with Latent Actions
Summary
This research proposes LAC-WM, a robot world model that operates in a unified latent action space shared across diverse robot embodiments. It outperforms explicit action-conditioned models in dexter, manipulation tasks and the LIBERO benchmark, demonstrating the benefits of a unified action space for cross-embodiment learning.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
Source：arXiv Robotics — research abstracts · arxiv.org