MobileVISTA: Generative Data Augmentation for Pose Generalization in Mobile Manipulation

Excerpt

arXiv:2610.07511v1 Announce Type: cross Abstract: Mobile manipulators such as humanoid robots are increasingly deployed in dynamic, unstructured environments to perform dexterous manipulation tasks. However, end-to-end manipulation policies trained to imitate demonstration data collected from a single robot pose are brittle: even centimeter-scale deviations in robot pose at deployment can drive ego-centric observations and end-effector trajectories out of the training distribution, leading to sh