Learning to Identify Critical States for Reinforcement Learning from Videos
This work addresses the challenge of leveraging unlabeled video data for reinforcement learning, offering a method to improve agent behavior understanding, though it appears incremental as it builds on prior offline DRL techniques.
The paper tackles the problem of extracting policy information from videos without action annotations by introducing Deep State Identifier, which predicts returns from video episodes and identifies critical states via mask-based sensitivity analysis, demonstrating its potential in experiments.
Recent work on deep reinforcement learning (DRL) has pointed out that algorithmic information about good policies can be extracted from offline data which lack explicit information about executed actions. For example, videos of humans or robots may convey a lot of implicit information about rewarding action sequences, but a DRL machine that wants to profit from watching such videos must first learn by itself to identify and recognize relevant states/actions/rewards. Without relying on ground-truth annotations, our new method called Deep State Identifier learns to predict returns from episodes encoded as videos. Then it uses a kind of mask-based sensitivity analysis to extract/identify important critical states. Extensive experiments showcase our method's potential for understanding and improving agent behavior. The source code and the generated datasets are available at https://github.com/AI-Initiative-KAUST/VideoRLCS.