IR3D-Bench: Evaluating Vision-Language Model Scene Understanding as Agentic Inverse Rendering

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