ResWAM:世界動作模型學習變化,而非保持不變的內容
ResWAM: Learning What Changes, Not What Stays, for World-Action Models
ResWAM以凍結的DINOv3特徵進行視覺殘差學習,並以0.5B雙分支DiT提升視覺語言動作策略成功率,同時降低訓練與推理計算量。模型總參數量為2.3B,在LIBERO-Plus、RoboTwin 2.0及RoboDojo分別取得81.1%、93.8%及9.40%的成功率,訓練計算量為2.7-4.0 TFLOPs,推理計算量為1.0-1.4 TFLOPs。
Published on Nov 14
Authors:
,
,
,
Abstract
ResWAM combines visual residual learning in frozen DINOv3 features with a lightweight dual-branch DiT trained via conditional flow matching to improve vision-language-action policy success while lowering training and inference compute.
World-action models (WAMs) augment vision-language-action policies (VLAs) with future prediction so that the policy can anticipate how its actions will affect the physical world. Existing WAMs supervise complete future observations. However, observations typically change little over an action chunk, so the future loss is dominated by copying static background content rather than learning task-relevant robot and object motion, limiting policy learning. Moreover, existing WAMs rely on large video generation models or separate world models, incurring heavy training and inference compute that hinders closed-loop deployment. To improve policy performance, we propose visual residual learning, supervising only the feature change from the current frame to the future frame in a frozen DINOv3 space. Compared with complete future features, this residual learning target yields a higher feature rank with richer and more diverse representations of scene changes, and spatially concentrates on the robot arm and manipulated objects rather than the background. To further improve computational efficiency, we present ResWAM, a residual world-action model with 2.3B parameters in total, including a lightweight 0.5B dual-branch DiT that jointly trains an action branch and a residual branch via conditional flow matching. Without robot pretraining (i.e., pretraining on external robot action trajectories beyond the benchmark demonstrations), ResWAM achieves top average success rates on LIBERO-Plus (81.1%), RoboTwin 2.0 (93.8%), and RoboDojo (9.40% success, 13.41 score) at the lowest training (2.7-4.0 TFLOPs) and inference compute (1.0-1.4 TFLOPs), and leads real-robot Open, Put, and Stack (80.0% average, +10.0% over π_{0.5}) at 92.7 ms latency.
View arXiv page View PDF Project page GitHub Add to collection
Models citing this paper 3
Robotics • Updated about 14 hours agoOPTML-Group/ResWAM-LIBERO
Robotics • Updated about 14 hours agoOPTML-Group/ResWAM-RoboTwin2
Robotics • Updated about 14 hours agoOPTML-Group/ResWAM-RoboDojo
Datasets citing this paper 0
No dataset linking this paper
Cite arxiv.org/abs/2611.19559 in a dataset README.md to link it from this page.
Spaces citing this paper 0
No Space linking this paper
Cite arxiv.org/abs/2611.19559 in a Space README.md to link it from this page.
Collections including this paper 1
來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co