TGM-VLA: Task-Guided Mixup for Sampling-Efficient and Robust Robotic Manipulation

返回详情 VLA / Vision-Language-Action 每日论文卡