RoboVIP: Multi-View Video Generation with Visual Identity Prompting Augments Robot Manipulation

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