SiGnature: Explicit Motion Diffusion for Stylized Semantic Gesture

2026-06-14

AI Summary

This work addresses the challenge of simultaneously achieving semantic accuracy and preserving speaker-specific nonverbal stylistic traits in co-speech gesture generation. To this end, we propose a diffusion-based framework that decouples semantic control from style retention in an explicit joint rotation space. Our approach introduces a plug-and-play Joint Motion Integration (JMI) mechanism that, without requiring retraining, seamlessly incorporates arbitrary real-world semantic motions during inference and mitigates stitching artifacts through external semantic injection. Experimental results demonstrate that the proposed method generates natural and fluent gestures while significantly improving both semantic motion control precision and speaker-style fidelity, outperforming current state-of-the-art baselines.

Abstract

While recent advances in co-speech gesture generation have achieved impressive rhythmic synchronization, synthesizing gestures that are both semantically meaningful and faithful to a speaker's unique non-verbal style remains an open challenge. Semantic gestures, such as iconic shapes or deictic pointing, are statistically sparse, making them difficult to learn effectively within standard generative models. We present SiGnature, a framework for Stylized and Semantic Gesture generation that reconciles precise semantic control with high-fidelity style preservation. Unlike prevalent methods that rely on entangled latent representations, SiGnature operates in an explicit joint-rotation space. This design enables our core contribution, Joint Motion Integration (JMI), a training-free inference mechanism capable of injecting any external motion sequence, particularly in-the-wild semantic gestures, directly into the diffusion process. JMI automatically identifies the specific ``active joints'' conveying a semantic action and injects them into the generation, while relying on the diffusion backbone to synthesize the remaining body dynamics, including posture and flow, in accordance with the pre-learned style of the target speaker. This allows for the plug-and-play integration of arbitrary motions, including complex semantic gestures, without retraining or introducing the ``Frankenstein'' artifacts typical of cut-and-paste methods. Extensive experiments and perceptual studies demonstrate that SiGnature offers superior semantic motion control while maintaining smooth and natural co-speech gesture generation and preserving the distinct characteristics of the speaker, thereby outperforming state-of-the-art baselines.

Problem

• co-speech gesture generation
• semantic gestures
• non-verbal style
• gesture synthesis
• stylized motion

Innovation

• explicit motion diffusion
• Joint Motion Integration
• stylized gesture generation
• semantic gesture control
• training-free inference