Online Prototype Alignment for Few-shot Policy Transfer
This addresses the challenge of efficient policy transfer in RL for scenarios with limited target data and visual differences, though it appears incremental as it builds on existing domain adaptation approaches.
The paper tackles the problem of domain adaptation in reinforcement learning when target domain data is scarce and visual similarity is unreliable, proposing Online Prototype Alignment (OPA) to achieve few-shot policy transfer within several episodes, with experimental results showing better performance and fewer samples needed compared to prior methods.
Domain adaptation in reinforcement learning (RL) mainly deals with the changes of observation when transferring the policy to a new environment. Many traditional approaches of domain adaptation in RL manage to learn a mapping function between the source and target domain in explicit or implicit ways. However, they typically require access to abundant data from the target domain. Besides, they often rely on visual clues to learn the mapping function and may fail when the source domain looks quite different from the target domain. To address these problems, we propose a novel framework Online Prototype Alignment (OPA) to learn the mapping function based on the functional similarity of elements and is able to achieve the few-shot policy transfer within only several episodes. The key insight of OPA is to introduce an exploration mechanism that can interact with the unseen elements of the target domain in an efficient and purposeful manner, and then connect them with the seen elements in the source domain according to their functionalities (instead of visual clues). Experimental results show that when the target domain looks visually different from the source domain, OPA can achieve better transfer performance even with much fewer samples from the target domain, outperforming prior methods.