DPoser: Diffusion Model as Robust 3D Human Pose Prior
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2023
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913279056543744 |
|---|---|
| author | Lu, Junzhe Lin, Jing Dou, Hongkun Zeng, Ailing Deng, Yue Zhang, Yulun Wang, Haoqian |
| author_facet | Lu, Junzhe Lin, Jing Dou, Hongkun Zeng, Ailing Deng, Yue Zhang, Yulun Wang, Haoqian |
| contents | This work targets to construct a robust human pose prior. However, it remains a persistent challenge due to biomechanical constraints and diverse human movements. Traditional priors like VAEs and NDFs often exhibit shortcomings in realism and generalization, notably with unseen noisy poses. To address these issues, we introduce DPoser, a robust and versatile human pose prior built upon diffusion models. DPoser regards various pose-centric tasks as inverse problems and employs variational diffusion sampling for efficient solving. Accordingly, designed with optimization frameworks, DPoser seamlessly benefits human mesh recovery, pose generation, pose completion, and motion denoising tasks. Furthermore, due to the disparity between the articulated poses and structured images, we propose truncated timestep scheduling to enhance the effectiveness of DPoser. Our approach demonstrates considerable enhancements over common uniform scheduling used in image domains, boasting improvements of 5.4%, 17.2%, and 3.8% across human mesh recovery, pose completion, and motion denoising, respectively. Comprehensive experiments demonstrate the superiority of DPoser over existing state-of-the-art pose priors across multiple tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2312_05541 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | DPoser: Diffusion Model as Robust 3D Human Pose Prior Lu, Junzhe Lin, Jing Dou, Hongkun Zeng, Ailing Deng, Yue Zhang, Yulun Wang, Haoqian Computer Vision and Pattern Recognition This work targets to construct a robust human pose prior. However, it remains a persistent challenge due to biomechanical constraints and diverse human movements. Traditional priors like VAEs and NDFs often exhibit shortcomings in realism and generalization, notably with unseen noisy poses. To address these issues, we introduce DPoser, a robust and versatile human pose prior built upon diffusion models. DPoser regards various pose-centric tasks as inverse problems and employs variational diffusion sampling for efficient solving. Accordingly, designed with optimization frameworks, DPoser seamlessly benefits human mesh recovery, pose generation, pose completion, and motion denoising tasks. Furthermore, due to the disparity between the articulated poses and structured images, we propose truncated timestep scheduling to enhance the effectiveness of DPoser. Our approach demonstrates considerable enhancements over common uniform scheduling used in image domains, boasting improvements of 5.4%, 17.2%, and 3.8% across human mesh recovery, pose completion, and motion denoising, respectively. Comprehensive experiments demonstrate the superiority of DPoser over existing state-of-the-art pose priors across multiple tasks. |
| title | DPoser: Diffusion Model as Robust 3D Human Pose Prior |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2312.05541 |