Learning Task Informed Abstractions
Abstract
Current model-based reinforcement learning methods struggle when operating from complex visual scenes due to their inability to prioritize task-relevant features. To mitigate this problem, we propose learning Task Informed Abstractions (TIA) that explicitly separates reward-correlated visual features from distractors.This approach is particularly useful in domains with rich visual observations where distinguishing between task-relevant and irrelevant information is crucial for sample efficiency. For learning TIA, we introduce the formalism of Task Informed MDP (TiMDP) that is realized by training two models that learn visual features via cooperative reconstruction, but one model is adversarially dissociated from the reward signal. Empirical evaluation shows that TIA leads to significant performance gains over state-of-the-art methods on many visual control tasks where natural and unconstrained visual distractions pose a formidable challenge.Our experiments were conducted on DMControl Suite with natural video backgrounds, which provides a challenging testbed for visual robustness.
Qualitative Results
