SMP: Reusable Score-Matching Motion Priors for Physics-Based Character Control
About
Data-driven motion priors that can guide agents toward producing naturalistic behaviors play a pivotal role in creating life-like virtual characters. Adversarial imitation learning has been a highly effective method for learning motion priors from reference motion data. However, adversarial priors, with few exceptions, need to be retrained for each new controller, thereby limiting their reusability and necessitating the retention of the reference motion data when training on downstream tasks. In this work, we present Score-Matching Motion Priors (SMP), which leverages pre-trained motion diffusion models and score distillation sampling (SDS) to create reusable task-agnostic motion priors. SMPs can be pre-trained on a motion dataset, independent of any control policy or task. Once trained, SMPs can be kept frozen and reused as general-purpose reward functions to train policies to produce naturalistic behaviors for downstream tasks. We show that a general motion prior trained on large-scale datasets can be repurposed into a variety of style-specific priors. Furthermore SMP can compose different styles to synthesize new styles not present in the original dataset. Our method produces high-quality motion comparable to state-of-the-art adversarial imitation learning methods through reusable and modular motion priors. We demonstrate the effectiveness of SMP across a diverse suite of control tasks with physically simulated humanoid characters. Video demo available at https://youtu.be/ravlZJteS20
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| open drawer | ANYTASK | Success Rate99.9 | 7 | |
| Lift Banana | ANYTASK | Success Rate0.00e+0 | 7 | |
| Lift Brick | ANYTASK | Success Rate0.00e+0 | 7 | |
| Lift Peach | ANYTASK | Success Rate0.00e+0 | 7 | |
| Push Pear to Center | ANYTASK | Success Rate2.1 | 7 | |
| Put Object In Closed Drawer | ANYTASK | Success Rate0.00e+0 | 7 | |
| Place Strawberry In Bowl | ANYTASK | Success Rate1.2 | 7 | |
| Stack Banana on Can | ANYTASK | Success Rate0.00e+0 | 7 |