DIDiffGes: Decoupled Semi-Implicit Diffusion Models for Real-time Gesture Generation from Speech
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908276906524672 |
|---|---|
| author | Cheng, Yongkang Huang, Shaoli Chen, Xuelin Ning, Jifeng Gong, Mingming |
| author_facet | Cheng, Yongkang Huang, Shaoli Chen, Xuelin Ning, Jifeng Gong, Mingming |
| contents | Diffusion models have demonstrated remarkable synthesis quality and diversity in generating co-speech gestures. However, the computationally intensive sampling steps associated with diffusion models hinder their practicality in real-world applications. Hence, we present DIDiffGes, for a Decoupled Semi-Implicit Diffusion model-based framework, that can synthesize high-quality, expressive gestures from speech using only a few sampling steps. Our approach leverages Generative Adversarial Networks (GANs) to enable large-step sampling for diffusion model. We decouple gesture data into body and hands distributions and further decompose them into marginal and conditional distributions. GANs model the marginal distribution implicitly, while L2 reconstruction loss learns the conditional distributions exciplictly. This strategy enhances GAN training stability and ensures expressiveness of generated full-body gestures. Our framework also learns to denoise root noise conditioned on local body representation, guaranteeing stability and realism. DIDiffGes can generate gestures from speech with just 10 sampling steps, without compromising quality and expressiveness, reducing the number of sampling steps by a factor of 100 compared to existing methods. Our user study reveals that our method outperforms state-of-the-art approaches in human likeness, appropriateness, and style correctness. Project is https://cyk990422.github.io/DIDiffGes. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_17059 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DIDiffGes: Decoupled Semi-Implicit Diffusion Models for Real-time Gesture Generation from Speech Cheng, Yongkang Huang, Shaoli Chen, Xuelin Ning, Jifeng Gong, Mingming Graphics Computer Vision and Pattern Recognition Sound Audio and Speech Processing Diffusion models have demonstrated remarkable synthesis quality and diversity in generating co-speech gestures. However, the computationally intensive sampling steps associated with diffusion models hinder their practicality in real-world applications. Hence, we present DIDiffGes, for a Decoupled Semi-Implicit Diffusion model-based framework, that can synthesize high-quality, expressive gestures from speech using only a few sampling steps. Our approach leverages Generative Adversarial Networks (GANs) to enable large-step sampling for diffusion model. We decouple gesture data into body and hands distributions and further decompose them into marginal and conditional distributions. GANs model the marginal distribution implicitly, while L2 reconstruction loss learns the conditional distributions exciplictly. This strategy enhances GAN training stability and ensures expressiveness of generated full-body gestures. Our framework also learns to denoise root noise conditioned on local body representation, guaranteeing stability and realism. DIDiffGes can generate gestures from speech with just 10 sampling steps, without compromising quality and expressiveness, reducing the number of sampling steps by a factor of 100 compared to existing methods. Our user study reveals that our method outperforms state-of-the-art approaches in human likeness, appropriateness, and style correctness. Project is https://cyk990422.github.io/DIDiffGes. |
| title | DIDiffGes: Decoupled Semi-Implicit Diffusion Models for Real-time Gesture Generation from Speech |
| topic | Graphics Computer Vision and Pattern Recognition Sound Audio and Speech Processing |
| url | https://arxiv.org/abs/2503.17059 |