Discovering Diverse Athletic Jumping Strategies

TL;DR

This paper introduces a framework combining physics simulation, deep reinforcement learning, and a pose autoencoder to automatically discover diverse, realistic athletic jumping strategies without relying on motion examples.

Contribution

It presents a novel approach that enables the emergence of diverse athletic motions through exploration and optimization, reducing the need for motion data and extensive reward engineering.

Findings

Discovered a variety of high jump strategies automatically.

Generated natural-looking athletic motions without motion examples.

Enhanced diversity of strategies through Bayesian search and optimization.

Abstract

We present a framework that enables the discovery of diverse and natural-looking motion strategies for athletic skills such as the high jump. The strategies are realized as control policies for physics-based characters. Given a task objective and an initial character configuration, the combination of physics simulation and deep reinforcement learning (DRL) provides a suitable starting point for automatic control policy training. To facilitate the learning of realistic human motions, we propose a Pose Variational Autoencoder (P-VAE) to constrain the actions to a subspace of natural poses. In contrast to motion imitation methods, a rich variety of novel strategies can naturally emerge by exploring initial character states through a sample-efficient Bayesian diversity search (BDS) algorithm. A second stage of optimization that encourages novel policies can further enrich the unique strategies discovered. Our method allows for the discovery of diverse and novel strategies for athletic jumping motions such as high jumps and obstacle jumps with no motion examples and less reward engineering than prior work.