Visual Reinforcement Learning with Imagined Goals
This work addresses the challenge of general-purpose skill acquisition for robotics, though it is incremental as it builds on existing goal-conditioned and representation learning approaches.
The paper tackles the problem of enabling autonomous agents to learn broadly applicable skills from raw sensory inputs like images, by combining unsupervised representation learning and reinforcement learning with imagined goals, resulting in a method that substantially outperforms prior techniques on real-world robotic systems.
For an autonomous agent to fulfill a wide range of user-specified goals at test time, it must be able to learn broadly applicable and general-purpose skill repertoires. Furthermore, to provide the requisite level of generality, these skills must handle raw sensory input such as images. In this paper, we propose an algorithm that acquires such general-purpose skills by combining unsupervised representation learning and reinforcement learning of goal-conditioned policies. Since the particular goals that might be required at test-time are not known in advance, the agent performs a self-supervised "practice" phase where it imagines goals and attempts to achieve them. We learn a visual representation with three distinct purposes: sampling goals for self-supervised practice, providing a structured transformation of raw sensory inputs, and computing a reward signal for goal reaching. We also propose a retroactive goal relabeling scheme to further improve the sample-efficiency of our method. Our off-policy algorithm is efficient enough to learn policies that operate on raw image observations and goals for a real-world robotic system, and substantially outperforms prior techniques.