PhysMoDPO: Physically-Plausible Humanoid Motion with Preference Optimization

Recent progress in text-conditioned human motion generation has been largely driven by diffusion models trained on large-scale human motion data. Building on this progress, recent methods attempt to transfer such models for character animation and real robot control by applying a Whole-Body Controller (WBC) that converts diffusion-generated motions into executable trajectories. While WBC trajectories become compliant with physics, they may expose substantial deviations from original motion. To address this issue, we here propose PhysMoDPO, a Direct Preference Optimization framework. Unlike prior work that relies on hand-crafted physics-aware heuristics such as foot-sliding penalties, we integrate WBC into our training pipeline and optimize diffusion model such that the output of WBC becomes compliant both with physics and original text instructions. To train PhysMoDPO we deploy physics-based and task-specific rewards and use them to assign preference to synthesized trajectories. Our extensive experiments on text-to-motion and spatial control tasks demonstrate consistent improvements of PhysMoDPO in both physical realism and task-related metrics on simulated robots. Moreover, we demonstrate that PhysMoDPO results in significant improvements when applied to zero-shot motion transfer in simulation and for real-world deployment on a G1 humanoid robot.

Learning feasible transitions for efficient contact planning

Contact planning for legged robots in extremely constrained environments is challenging. The main difficulty stems from the mixed nature of the problem, discrete search together with continuous trajectory optimization. To speed up the discrete search problem, we propose in this paper to learn the properties of transitions from one contact mode to the next. In particular, we learn a feasibility classifier and an offset network; the former predicts if a potential next contact state is feasible from the current contact state, while the latter learns to compensate for misalignment in achieving a desired contact state due to imperfections of the low-level control. We integrate these learned networks in a Monte Carlo Tree Search (MCTS) contact planner to better prune the tree and improve the heuristic. Our simulation results demonstrate that training these networks with offline data significantly speeds up the online search process and improves its accuracy.