Learning Actionable Representations from Visual Observations
In this work, we explore a new approach for robots to teach themselves about the world simply by observing it. In particular, we investigate the effectiveness of learning task-agnostic representations for continuous control tasks. We extend Time-Contrastive Networks (TCN) that learn from visual observations by embedding multiple frames jointly in the embedding space as opposed to a single frame. We show that by doing so, we are now able to encode both position and velocity attributes significantly more accurately. We test the usefulness of this self-supervised approach in a reinforcement learning setting. We show that the representations learned by agents observing themselves take random actions, or other agents perform tasks successfully, can enable the learning of continuous control policies using algorithms like Proximal Policy Optimization (PPO) using only the learned embeddings as input. We also demonstrate significant improvements on the real-world Pouring dataset with a relative error reduction of 39.4% for motion attributes and 11.1% for static attributes compared to the single-frame baseline.
Step 1: Observe agents interacting with their environments from multiple views
Cartpole taking random actions
Cheetah performing Walk task
Step 2: Learn visual representations by embedding multiple frames jointly
Step 3: Train control policies with learned visual representations as input