Run, skeleton, run: skeletal model in a physics-based simulation
This work addresses the problem of efficient and stable training in high-dimensional, stochastic physics simulations for researchers in reinforcement learning and robotics, though it is incremental with benchmarked methods and minor improvements.
The paper tackled training a physiologically-based human model to navigate a complex obstacle course quickly in a physics-based reinforcement learning challenge, finding that Deep Deterministic Policy Gradient with improvements like layer normalization stabilized training and allowed generalization to new scenarios.
In this paper, we present our approach to solve a physics-based reinforcement learning challenge "Learning to Run" with objective to train physiologically-based human model to navigate a complex obstacle course as quickly as possible. The environment is computationally expensive, has a high-dimensional continuous action space and is stochastic. We benchmark state of the art policy-gradient methods and test several improvements, such as layer normalization, parameter noise, action and state reflecting, to stabilize training and improve its sample-efficiency. We found that the Deep Deterministic Policy Gradient method is the most efficient method for this environment and the improvements we have introduced help to stabilize training. Learned models are able to generalize to new physical scenarios, e.g. different obstacle courses.