Mirage2Matter: A Physically Grounded Gaussian World Model from Video

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