LaMP: Learning Vision-Language-Action Policies with 3D Scene Flow as Latent Motion Prior

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