DriveMA: Rethinking Language Interfaces in Driving VLAs with One-Step Meta-Actions

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