NaP-Control: Navigating Diffusion Prior for Versatile and Fast Character Control

TL;DR

NaP-Control leverages reinforcement learning to steer diffusion priors for fast, robust, and versatile character control in physics-based animation, outperforming existing methods in success rate and efficiency.

Contribution

The paper introduces NaP-Control, a novel approach that manipulates diffusion prior noise via reinforcement learning for improved, fast, and adaptable character control.

Findings

NaP achieves higher success rates across diverse tasks.

NaP enables faster inference compared to gradient-based guidance.

NaP maintains natural motion quality in complex scenarios.

Abstract

Achieving precise, versatile whole-body character control in physics-based animation remains challenging. Recent diffusion-based policies generate rich and expressive motions but typically rely on gradient-based test-time guidance to satisfy task objectives, which is slow and can reduce robustness. We introduce NaP-Control (Navigating Diffusion Prior for Versatile and Fast Character Control), abbreviated as NaP. Our method uses reinforcement learning to manipulate the latent noise of a task-agnostic diffusion policy prior, steering it toward task-specific behaviors for fast, robust control with high motion fidelity. In contrast to methods that rely solely on offline training, NaP interacts with the environment during training to correct motions and optimize task rewards, improving success rates and enabling adaptation to challenging scenarios. By directly predicting task-optimized diffusion noise, NaP eliminates iterative guidance during denoising and enables efficient inference. Experiments show that NaP attains higher success rates and faster inference while preserving natural motion across diverse tasks.