Walk the Random Walk: Learning to Discover and Reach Goals Without Supervision
This addresses the challenge of unsupervised skill discovery for applications like robotics, though it is incremental as it builds on existing goal-conditioned learning methods.
The paper tackles the problem of training a goal-conditioned agent to reach diverse states without external supervision, achieving this by using a random walk to train a reachability network and goal memory, with results demonstrated on continuous control navigation and robotic manipulation tasks.
Learning a diverse set of skills by interacting with an environment without any external supervision is an important challenge. In particular, obtaining a goal-conditioned agent that can reach any given state is useful in many applications. We propose a novel method for training such a goal-conditioned agent without any external rewards or any domain knowledge. We use random walk to train a reachability network that predicts the similarity between two states. This reachability network is then used in building goal memory containing past observations that are diverse and well-balanced. Finally, we train a goal-conditioned policy network with goals sampled from the goal memory and reward it by the reachability network and the goal memory. All the components are kept updated throughout training as the agent discovers and learns new goals. We apply our method to a continuous control navigation and robotic manipulation tasks.