Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Task-Relevant Representation Decoupling for Visual Reinforcement Learning Generalization

About

Visual Reinforcement Learning (VRL) has achieved considerable success in solving control tasks. However, generalizing learned policies to new environments remains a major challenge, as agents often overfit to task-irrelevant features in the training environment. To solve this problem, we introduce the concept of decoupling observations into task-relevant and task-irrelevant representations. Building on this idea, we propose a self-supervised Task-Relevant Representation Decoupling (T2RD) algorithm for VRL. This algorithm consists of three components: task-relevant representation consistency, cross-reconstruction, and cross-dynamic prediction. The first two components achieve the decoupling of content and style features, but the resulting content representations are not necessarily task-relevant. To further refine task-relevant features from content representations, we design the third component that introduces dynamic prediction. T2RD achieves State-Of-The-Art (SOTA) generalization performance and sample efficiency in the DeepMind Control Suite and Robotic Manipulation tasks.

Jinwen Wang, Youfang Lin, Xiaobo Hu, Qian Xu, Shuo Wang, Zhuo Chen, Kai Lv• 2026

Related benchmarks

TaskDatasetResultRank
Reinforcement LearningDMC-GB2 Video Hard (test)
Cartpole Swingup Return725
15
Robotic Manipulationpeg-in-box (test2)
Return129.4
14
Visual Reinforcement LearningDMC-GB Color Hard
Average Return: Walker, Walk849
10
Visual Reinforcement LearningDMControl-GB Video-Easy
Walker Walk Score821
10
Peg in boxRobotic Manipulation (Test1)
Episode Return164.1
7
Peg in boxRobotic Manipulation (Test3)
Episode Return183.6
7
Peg in boxRobotic Manipulation (Test4)
Episode Return149.8
7
ReachRobotic Manipulation (Test1)
Episode Return25.1
7
ReachRobotic Manipulation (Test3)
Episode Return29.8
7
ReachRobotic Manipulation (Test4)
Episode Return28.3
7
Showing 10 of 11 rows

Other info

Follow for update