T-Rex: Tactile-Reactive Dexterous Manipulation
触觉+vla反应式控制，预训练+中训练+后训练，预训练不断沉淀更通用的物理与运动先验，后训练不断变轻、变快。触觉模态的加入，正是在把 “接触物理” 这部分也纳入预训练的通用知识中，逐步缩小后训练的必要性。
2 Matching Annotations
- Last 7 days
-
-
-
arxiv.org arxiv.org
-
Figure 1: Overview of X-WAM. Top: X-WAM is a unified 4D World Action Model that jointly predicts future multi-view RGB-D videos and robot actions from video priors, featuring a lightweight depth adaptation module for spatial reconstruction and Asynchronous Noise Sampling (ANS) for efficient action decoding. Bottom: X-WAM surpasses existing methods in policy success rate on RoboCasa and RoboTwin 2.0, produces high-fidelity 4D reconstruction and generation, and enables real-time execution deployment on physical robots.
联合生成4D视频结果以及机器人的未来状态与动作
-