arXiv:2610.37518 · v1

World Dreamer v5-lite: A Multi-Robot World Action Model with One Dreamer

首次提交 2026-10-28 00:40

版本更新 2026-10-28 00:40

雷达观测 2026-10-29 00:34

摘要

Multi-robot control requires integrating distributed camera views while predicting separate action sequences. Most vision-language-action and world-action models target one robot. Combining multiple views alone does not establish spatial correspondence between them. World Dreamer v5-lite is a multi-robot world-action model built around a shared world state for robot-specific action prediction. Calibrated features from a pretrained video encoder populate this state in workspace coordinates; its spatial size remains fixed as views are added. A vision-language model provides scene context and one query per robot. The queries read the shared state and exchange pose-aware information, while their action streams remain separate. Masked-view cooperative training withholds selected camera views from both visual branches while retaining all action targets. This encourages the use of observations shared by other robots. Pretraining uses more than 10,000 hours of trajectories from 20 simulated robot types, RoboFactory, and real robots. On RoboFactory, World Dreamer v5-lite achieves 74.0% mean success, 23.7 percentage points above the strongest evaluated baseline. Across four real-robot tasks, it reaches 78.8% mean success, a margin of 13.8 points. It also reaches 58.08% overall on RoboCasa365.

作者备注：10 pages, 4 figures. Guoning Wu and Zijian Cai share first authorship. Yongbin Zheng is the corresponding author