Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.14161 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917212161310720 |
|---|---|
| author | Dong, Yitong Zhang, Qi Jiang, Minchao Wu, Zhiqiang Fan, Qingnan Feng, Ying Zhang, Huaqi Bao, Hujun Zhang, Guofeng |
| author_facet | Dong, Yitong Zhang, Qi Jiang, Minchao Wu, Zhiqiang Fan, Qingnan Feng, Ying Zhang, Huaqi Bao, Hujun Zhang, Guofeng |
| contents | We present a novel framework for high-fidelity novel view synthesis (NVS) from sparse images, addressing key limitations in recent feed-forward 3D Gaussian Splatting (3DGS) methods built on Vision Transformer (ViT) backbones. While ViT-based pipelines offer strong geometric priors, they are often constrained by low-resolution inputs due to computational costs. Moreover, existing generative enhancement methods tend to be 3D-agnostic, resulting in inconsistent structures across views, especially in unseen regions. To overcome these challenges, we design a Dual-Domain Detail Perception Module, which enables handling high-resolution images without being limited by the ViT backbone, and endows Gaussians with additional features to store high-frequency details. We develop a feature-guided diffusion network, which can preserve high-frequency details during the restoration process. We introduce a unified training strategy that enables joint optimization of the ViT-based geometric backbone and the diffusion-based refinement module. Experiments demonstrate that our method can maintain superior generation quality across multiple datasets. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_14161 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | One-Shot Refiner: Boosting Feed-forward Novel View Synthesis via One-Step Diffusion Dong, Yitong Zhang, Qi Jiang, Minchao Wu, Zhiqiang Fan, Qingnan Feng, Ying Zhang, Huaqi Bao, Hujun Zhang, Guofeng Computer Vision and Pattern Recognition We present a novel framework for high-fidelity novel view synthesis (NVS) from sparse images, addressing key limitations in recent feed-forward 3D Gaussian Splatting (3DGS) methods built on Vision Transformer (ViT) backbones. While ViT-based pipelines offer strong geometric priors, they are often constrained by low-resolution inputs due to computational costs. Moreover, existing generative enhancement methods tend to be 3D-agnostic, resulting in inconsistent structures across views, especially in unseen regions. To overcome these challenges, we design a Dual-Domain Detail Perception Module, which enables handling high-resolution images without being limited by the ViT backbone, and endows Gaussians with additional features to store high-frequency details. We develop a feature-guided diffusion network, which can preserve high-frequency details during the restoration process. We introduce a unified training strategy that enables joint optimization of the ViT-based geometric backbone and the diffusion-based refinement module. Experiments demonstrate that our method can maintain superior generation quality across multiple datasets. |
| title | One-Shot Refiner: Boosting Feed-forward Novel View Synthesis via One-Step Diffusion |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2601.14161 |