CVPR 2026研究：视频世界模型无法可靠控制机器人
CVPR最新研究揭示视频模型与机器人控制的鸿沟，GEM-4D如何提升操作成功率
CVPR 2026 WMAS研讨会上，最佳论文RoboWM-Bench发现视频世界模型能生成逼真画面，但预测行为转为机器人动作时失败。失败源于空间推理不稳定和接触问题。GEM-4D通过添加帧间几何一致性，将现实操作成功率从61%提升至81%。SAW-Bench测试显示，人类与最佳多模态模型在情境感知任务上存在37.66分差距。
A realistic video is not enough for a robot to act reliably.
That was the central question at CVPR 2026’s WMAS workshop. RoboWM-Bench, the Best Paper, found that a video world model can produce convincing footage while still failing when its predicted behavior is turned into robot actions. The failures came from problems such as spatial reasoning and unstable contact. GEM-4D improved real-world manipulation success from 61% to 81% by adding geometric consistency across generated frames. SAW-Bench found a 37.66-point gap between humans and the best multimodal model on situated-awareness tasks.
https://t.co/Dd3xPaok2w
