BRIC: Bridging Kinematic Plans and Physical Control at Test Time

TL;DR

BRIC is a test-time framework that improves long-term human motion generation by adapting physics controllers and guiding diffusion models, ensuring physically plausible and consistent motions in diverse environments.

Contribution

BRIC introduces a novel combination of dynamic physics controller adaptation and diffusion model guidance for improved long-term motion generation.

Findings

Achieves state-of-the-art performance on multiple long-term motion tasks.

Ensures physically plausible and consistent motions across diverse environments.

Effectively combines adaptation strategies without retraining models.

Abstract

We propose BRIC, a novel test-time adaptation (TTA) framework that enables long-term human motion generation by resolving execution discrepancies between diffusion-based kinematic motion planners and reinforcement learning-based physics controllers. While diffusion models can generate diverse and expressive motions conditioned on text and scene context, they often produce physically implausible outputs, leading to execution drift during simulation. To address this, BRIC dynamically adapts the physics controller to noisy motion plans at test time, while preserving pre-trained skills via a loss function that mitigates catastrophic forgetting. In addition, BRIC introduces a lightweight test-time guidance mechanism that steers the diffusion model in the signal space without updating its parameters. By combining both adaptation strategies, BRIC ensures consistent and physically plausible long-term executions across diverse environments in an effective and efficient manner. We validate the effectiveness of BRIC on a variety of long-term tasks, including motion composition, obstacle avoidance, and human-scene interaction, achieving state-of-the-art performance across all tasks.