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