DenoiseSplat: Feed-Forward Gaussian Splatting for Noisy 3D Scene Reconstruction
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866912958412488704 |
|---|---|
| author | Jiang, Fuzhen Li, Zhuoran Zhang, Yinlin |
| author_facet | Jiang, Fuzhen Li, Zhuoran Zhang, Yinlin |
| contents | 3D scene reconstruction and novel-view synthesis are fundamental for VR, robotics, and content creation. However, most NeRF and 3D Gaussian Splatting pipelines assume clean inputs and degrade under real noise and artifacts. We therefore propose DenoiseSplat, a feed-forward 3D Gaussian splatting method for noisy multi-view images. We build a large-scale, scene-consistent noisy--clean benchmark on RE10K by injecting Gaussian, Poisson, speckle, and salt-and-pepper noise with controlled intensities. With a lightweight MVSplat-style feed-forward backbone, we train end-to-end using only clean 2D renderings as supervision and no 3D ground truth. On noisy RE10K, DenoiseSplat outperforms vanilla MVSplat and a strong two-stage baseline (IDF + MVSplat) in PSNR/SSIM and LPIPS across noise types and levels. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_09291 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | DenoiseSplat: Feed-Forward Gaussian Splatting for Noisy 3D Scene Reconstruction Jiang, Fuzhen Li, Zhuoran Zhang, Yinlin Computer Vision and Pattern Recognition Artificial Intelligence 3D scene reconstruction and novel-view synthesis are fundamental for VR, robotics, and content creation. However, most NeRF and 3D Gaussian Splatting pipelines assume clean inputs and degrade under real noise and artifacts. We therefore propose DenoiseSplat, a feed-forward 3D Gaussian splatting method for noisy multi-view images. We build a large-scale, scene-consistent noisy--clean benchmark on RE10K by injecting Gaussian, Poisson, speckle, and salt-and-pepper noise with controlled intensities. With a lightweight MVSplat-style feed-forward backbone, we train end-to-end using only clean 2D renderings as supervision and no 3D ground truth. On noisy RE10K, DenoiseSplat outperforms vanilla MVSplat and a strong two-stage baseline (IDF + MVSplat) in PSNR/SSIM and LPIPS across noise types and levels. |
| title | DenoiseSplat: Feed-Forward Gaussian Splatting for Noisy 3D Scene Reconstruction |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2603.09291 |