Neural probabilistic motor primitives for humanoid control
This addresses the challenge of flexible and robust humanoid control for robotics applications, representing an incremental advance in motor primitive learning.
The paper tackles the problem of learning a single motor module for controlling high-dimensional simulated humanoids by proposing a neural probabilistic motor primitive system trained offline to compress thousands of expert policies, enabling one-shot imitation of unseen whole-body behaviors and naturalistic task-solving movements.
We focus on the problem of learning a single motor module that can flexibly express a range of behaviors for the control of high-dimensional physically simulated humanoids. To do this, we propose a motor architecture that has the general structure of an inverse model with a latent-variable bottleneck. We show that it is possible to train this model entirely offline to compress thousands of expert policies and learn a motor primitive embedding space. The trained neural probabilistic motor primitive system can perform one-shot imitation of whole-body humanoid behaviors, robustly mimicking unseen trajectories. Additionally, we demonstrate that it is also straightforward to train controllers to reuse the learned motor primitive space to solve tasks, and the resulting movements are relatively naturalistic. To support the training of our model, we compare two approaches for offline policy cloning, including an experience efficient method which we call linear feedback policy cloning. We encourage readers to view a supplementary video ( https://youtu.be/CaDEf-QcKwA ) summarizing our results.