Abstract:How to improve the ability of scene representation is a key issue in vision-oriented decision-making applications, and current approaches usually learn task-relevant state representations within visual reinforcement learning to address this problem. While prior work typically introduces one-step behavioral similarity metrics with elements (e.g., rewards and actions) to extract task-relevant state information from observations, they often ignore the inherent dynamics relationships among the elements that are essential for learning accurate representations, which further impedes the discrimination of short-term similar task/behavior information in long-term dynamics transitions. To alleviate this problem, we propose an intrinsic dynamics-driven representation learning method with sequence models in visual reinforcement learning, namely DSR. Concretely, DSR optimizes the parameterized encoder by the state-transition dynamics of the underlying system, which prompts the latent encoding information to satisfy the state-transition process and then the state space and the noise space can be distinguished. In the implementation and to further improve the representation ability of DSR on encoding similar tasks, sequential elements' frequency domain and multi-step prediction are adopted for sequentially modeling the inherent dynamics. Finally, experimental results show that DSR has achieved significant performance improvements in the visual Distracting DMControl control tasks, especially with an average of 78.9\% over the backbone baseline. Further results indicate that it also achieves the best performances in real-world autonomous driving applications on the CARLA simulator. Moreover, qualitative analysis results validate that our method possesses the superior ability to learn generalizable scene representations on visual tasks. The source code is available at <a class="link-external link-https" href="https://github.com/DMU-XMU/DSR" rel="external noopener nofollow">this https URL</a>.

State Chrono Representation for Enhancing Generalization in Reinforcement Learning

MERIT: Learning Multi-level Representations on Temporal Graphs.

SimSR: Simple Distance-based State Representation for Deep Reinforcement Learning

Learning Invariant Representations for Reinforcement Learning without Reconstruction

Intrinsic Dynamics-Driven Generalizable Scene Representations for Vision-Oriented Decision-Making Applications

Policy-Independent Behavioral Metric-Based Representation for Deep Reinforcement Learning

Learning State Representations via Retracing in Reinforcement Learning

Learning a World Model With Multitimescale Memory Augmentation

Episodic Reinforcement Learning with Expanded State-reward Space

Prototypical context-aware dynamics generalization for high-dimensional model-based reinforcement learning

State Representation Learning for Effective Deep Reinforcement Learning.

Mix-up Consistent Cross Representations for Data-Efficient Reinforcement Learning

Bridging State and History Representations: Understanding Self-Predictive RL

Combining long and short spatiotemporal reasoning for deep reinforcement learning

Consciousness-Inspired Spatio-Temporal Abstractions for Better Generalization in Reinforcement Learning

Towards Generalizable Reinforcement Learning via Causality-Guided Self-Adaptive Representations

DynaMITE-RL: A Dynamic Model for Improved Temporal Meta-Reinforcement Learning

Domain Adaptive State Representation Alignment for Reinforcement Learning

Model-Free Generative Replay for Lifelong Reinforcement Learning: Application to Starcraft-2

Simoun: Synergizing Interactive Motion-appearance Understanding for Vision-based Reinforcement Learning

State Representation Learning with Adjacent State Consistency Loss for Deep Reinforcement Learning.