METIS: Multi-Source Egocentric Training for Integrated Dexterous Vision-Language-Action Model

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