Generating a Terrain-Robustness Benchmark for Legged Locomotion: A Prototype via Terrain Authoring and Active Learning
This work addresses a bottleneck for robotics researchers by providing a benchmark to improve terrain-aware locomotion, though it is incremental as it builds on existing terrain generation techniques.
The paper tackles the difficulty of generating diverse and realistic unstructured terrains in simulation for evaluating legged locomotion policies by prototyping a dataset creation method using terrain authoring and active learning, resulting in stable generation of high-quality terrains.
Terrain-aware locomotion has become an emerging topic in legged robotics. However, it is hard to generate diverse, challenging, and realistic unstructured terrains in simulation, which limits the way researchers evaluate their locomotion policies. In this paper, we prototype the generation of a terrain dataset via terrain authoring and active learning, and the learned samplers can stably generate diverse high-quality terrains. We expect the generated dataset to make a terrain-robustness benchmark for legged locomotion. The dataset, the code implementation, and some policy evaluations are released at https://bit.ly/3bn4j7f.