Run, skeleton, run: skeletal model in a physics-based simulation

TL;DR

This paper develops a physics-based reinforcement learning approach to train a human skeletal model to navigate obstacle courses efficiently, demonstrating improved stability and generalization across scenarios.

Contribution

The paper benchmarks policy-gradient methods for complex physics-based tasks and introduces stabilization techniques, with Deep Deterministic Policy Gradient proving most effective.

Findings

Deep Deterministic Policy Gradient outperforms other methods

Training stabilization techniques improve sample efficiency

Models generalize to new physical obstacle scenarios

Abstract

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.