UNIVERSE: Unified Video Action Models for Autonomous Driving with Flexible Mask-Modulated Modality Generation figure
AlphaXiv 中文概览(可滚动查看)