Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2509.05582 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915498884595712 |
|---|---|
| author | Ye, Zhiling Zhou, Cong Zhang, Xiubao Shen, Haifeng Deng, Weihong Lu, Quan |
| author_facet | Ye, Zhiling Zhou, Cong Zhang, Xiubao Shen, Haifeng Deng, Weihong Lu, Quan |
| contents | In this paper, we explore a reconstruction and reenactment separated framework for 3D Gaussians head, which requires only a single portrait image as input to generate controllable avatar. Specifically, we developed a large-scale one-shot gaussian head generator built upon WebSSL and employed a two-stage training approach that significantly enhances the capabilities of generalization and high-frequency texture reconstruction. During inference, an ultra-lightweight gaussian avatar driven by control signals enables high frame-rate rendering, achieving 90 FPS at a resolution of 512x512. We further demonstrate that the proposed framework follows the scaling law, whereby increasing the parameter scale of the reconstruction module leads to improved performance. Moreover, thanks to the separation design, driving efficiency remains unaffected. Finally, extensive quantitative and qualitative experiments validate that our approach outperforms current state-of-the-art methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_05582 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Reconstruction and Reenactment Separated Method for Realistic Gaussian Head Ye, Zhiling Zhou, Cong Zhang, Xiubao Shen, Haifeng Deng, Weihong Lu, Quan Computer Vision and Pattern Recognition In this paper, we explore a reconstruction and reenactment separated framework for 3D Gaussians head, which requires only a single portrait image as input to generate controllable avatar. Specifically, we developed a large-scale one-shot gaussian head generator built upon WebSSL and employed a two-stage training approach that significantly enhances the capabilities of generalization and high-frequency texture reconstruction. During inference, an ultra-lightweight gaussian avatar driven by control signals enables high frame-rate rendering, achieving 90 FPS at a resolution of 512x512. We further demonstrate that the proposed framework follows the scaling law, whereby increasing the parameter scale of the reconstruction module leads to improved performance. Moreover, thanks to the separation design, driving efficiency remains unaffected. Finally, extensive quantitative and qualitative experiments validate that our approach outperforms current state-of-the-art methods. |
| title | Reconstruction and Reenactment Separated Method for Realistic Gaussian Head |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2509.05582 |