EgoVLA: Learning Vision-Language-Action Models from Egocentric Human Videos figure
AlphaXiv 中文概览(可滚动查看)