Future Prediction as Training-Only Video Supervision for Pretrained VLA Policies
Hanwen Cui, Yuanyuan Li, Qing Guo, Zhida Zhang, Chao Chen, Xi Chen, Jie Cao
原文摘要
World action models (WAMs) introduce predictive supervision into robot policy learning by jointly modeling future visual observations and actions. Pretrained vision-language-action (VLA) policies acquire rich visual and semantic representations as well as action-generation capabilities. Their action objectives typically lack direct supervision for predicting future visual observations. Future videos in robot demonstrations record object motion and task progress. They can provide temporal visual supervision that complements action labels. We propose Training-Only Video Supervision, which introduces future-video prediction as an auxiliary objective for VLA fine-tuning. We condition a randomly initialized video-prediction expert on the VLA's shared vision-language features. The expert learns to predict future video latents from robot demonstrations. The video prediction loss and the original action loss jointly update the shared vision-language backbone, providing additional temporal visual supervision. After training, we remove the video expert and its adapter. The fine-tuned VLA retains its original action-generation pathway at inference. On the 50-task RoboTwin 2.0 benchmark, our method achieves 71.18% average success, compared with $π_{0.5}$ of 58.33%. In a matched evaluation on four real-world manipulation tasks, success increases from 42.50% to 61.25%. These results show that future visual modeling can serve as auxiliary training supervision for pretrained VLAs. The fine-tuned action policy benefits from this supervision without retaining a video-prediction branch at deployment.
cs.RO