Saved in:
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.01479 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911564204867584 |
|---|---|
| author | Huang, Zhisheng Chen, Jiahao Lin, Cheng Hu, Chenyu Huang, Hanzhuo Yu, Zhengming Li, Mengfei Liu, Yuheng Gu, Zekai Zhao, Zibo Liu, Yuan Li, Xin Wang, Wenping |
| author_facet | Huang, Zhisheng Chen, Jiahao Lin, Cheng Hu, Chenyu Huang, Hanzhuo Yu, Zhengming Li, Mengfei Liu, Yuheng Gu, Zekai Zhao, Zibo Liu, Yuan Li, Xin Wang, Wenping |
| contents | Sparse-view 3D modeling represents a fundamental tension between reconstruction fidelity and generative plausibility. While feed-forward reconstruction excels in efficiency and input alignment, it often lacks the global priors needed for structural completeness. Conversely, diffusion-based generation provides rich geometric details but struggles with multi-view consistency. We present UniRecGen, a unified framework that integrates these two paradigms into a single cooperative system. To overcome inherent conflicts in coordinate spaces, 3D representations, and training objectives, we align both models within a shared canonical space. We employ disentangled cooperative learning, which maintains stable training while enabling seamless collaboration during inference. Specifically, the reconstruction module is adapted to provide canonical geometric anchors, while the diffusion generator leverages latent-augmented conditioning to refine and complete the geometric structure. Experimental results demonstrate that UniRecGen achieves superior fidelity and robustness, outperforming existing methods in creating complete and consistent 3D models from sparse observations. Code is available at https://github.com/zsh523/UniRecGen. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_01479 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | UniRecGen: Unifying Multi-View 3D Reconstruction and Generation Huang, Zhisheng Chen, Jiahao Lin, Cheng Hu, Chenyu Huang, Hanzhuo Yu, Zhengming Li, Mengfei Liu, Yuheng Gu, Zekai Zhao, Zibo Liu, Yuan Li, Xin Wang, Wenping Computer Vision and Pattern Recognition Sparse-view 3D modeling represents a fundamental tension between reconstruction fidelity and generative plausibility. While feed-forward reconstruction excels in efficiency and input alignment, it often lacks the global priors needed for structural completeness. Conversely, diffusion-based generation provides rich geometric details but struggles with multi-view consistency. We present UniRecGen, a unified framework that integrates these two paradigms into a single cooperative system. To overcome inherent conflicts in coordinate spaces, 3D representations, and training objectives, we align both models within a shared canonical space. We employ disentangled cooperative learning, which maintains stable training while enabling seamless collaboration during inference. Specifically, the reconstruction module is adapted to provide canonical geometric anchors, while the diffusion generator leverages latent-augmented conditioning to refine and complete the geometric structure. Experimental results demonstrate that UniRecGen achieves superior fidelity and robustness, outperforming existing methods in creating complete and consistent 3D models from sparse observations. Code is available at https://github.com/zsh523/UniRecGen. |
| title | UniRecGen: Unifying Multi-View 3D Reconstruction and Generation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2604.01479 |