ResDiffFRG: Residual Diffusion for Multiple Appropriate Facial Reaction Generation

2026-09-27

AI Summary

This study addresses the challenge that existing diffusion models for generating multimodal facial reactions often produce complex denoising trajectories prone to deviating from plausible ranges due to random initialization. To overcome this, we propose ResDiffFRG, a novel framework that introduces the first speaker-anchored residual diffusion mechanism. By leveraging the behavioral mimicry prior between speakers and listeners, the framework reformulates the diffusion target as residuals relative to a speaker anchor, thereby significantly simplifying the denoising process. Experimental results demonstrate that our approach substantially outperforms baselines on correlation metrics, surpassing the performance of Gaussian diffusion after 60% of the denoising steps right from initialization. These findings confirm that the proposed method effectively enhances both the diversity and plausibility of the generated facial reactions.

Abstract

In dyadic human speaker-listener conversations, the listener's facial reactions allows the speaker to accurately perceive the listener's emotional states. Since human facial reactions are non-deterministic, the ability to generate multiple appropriate human-like facial reactions is crucial for realistic human-agent interactions. Although diffusion models are naturally suited to such one-to-many generation, existing diffusion-based Multiple Appropriate Facial Reaction Generation (MAFRG) methods attempt to denoise random Gaussian initialisations directly into multiple appropriate facial reactions (AFRs). These random initialisations are usually not well-aligned with the target listener facial reaction, which requires complex denoising trajectories from these initialisations, and subsequently creates substantial opportunities for deviations away from the range of trajectories leading to appropriate AFRs. Given the inherent mimicry between the human listener's and speaker's facial behaviours, we address the above denoising trajectory issue by leveraging this strong prior. Specifically, we propose ResDiffFRG, a novel diffusion-based MAFRG framework that explicitly anchors the diffusion process to the speaker behaviour by defining its diffusion target as the residual between the speaker anchor and an AFR. The denoiser only needs to model the comparatively small, reaction-specific residual needed to transform this anchor into an AFR, rather than reconstructing the complete reaction from an unstructured state. Extensive experiments show that ResDiffFRG achieves large improvements in correlation-based appropriateness over existing methods. Our denoising trajectory analysis showed that even at the start of the denoising trajectory, ResDiffFRG already achieves a higher facial-reaction correlation score than the Gaussian Diffusion baseline does after completing 60% of its denoising trajectory.

Problem

• Multiple Appropriate Facial Reaction Generation
• Diffusion Models
• Denoising Trajectory
• Human-Agent Interaction

Innovation

• Residual Diffusion
• Facial Reaction Generation
• One-to-Many Generation
• Denoising Trajectory
• Speaker Anchor