Bootstrap-GS: Self-Supervised Augmentation for High-Fidelity Gaussian Splatting
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909523142246400 |
|---|---|
| author | Gao, Yifei Ren, Kerui Ou, Jie Wang, Lei Wu, Jiaji Cheng, Jun |
| author_facet | Gao, Yifei Ren, Kerui Ou, Jie Wang, Lei Wu, Jiaji Cheng, Jun |
| contents | Recent advancements in 3D Gaussian Splatting (3D-GS) have established new benchmarks for rendering quality and efficiency in 3D reconstruction. However, 3D-GS faces critical limitations when generating novel views that significantly deviate from those encountered during training. Moreover, issues such as dilation and aliasing arise during zoom operations. These challenges stem from a fundamental issue: training sampling deficiency. In this paper, we introduce a bootstrapping framework to address this problem. Our approach synthesizes pseudo-ground truth from novel views that align with the limited training set and reintegrates these synthesized views into the training pipeline. Experimental results demonstrate that our bootstrapping technique not only reduces artifacts but also improves quantitative metrics. Furthermore, our technique is highly adaptable, allowing various Gaussian-based method to benefit from its integration. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_18669 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Bootstrap-GS: Self-Supervised Augmentation for High-Fidelity Gaussian Splatting Gao, Yifei Ren, Kerui Ou, Jie Wang, Lei Wu, Jiaji Cheng, Jun Graphics Artificial Intelligence Computer Vision and Pattern Recognition I.4.8 Recent advancements in 3D Gaussian Splatting (3D-GS) have established new benchmarks for rendering quality and efficiency in 3D reconstruction. However, 3D-GS faces critical limitations when generating novel views that significantly deviate from those encountered during training. Moreover, issues such as dilation and aliasing arise during zoom operations. These challenges stem from a fundamental issue: training sampling deficiency. In this paper, we introduce a bootstrapping framework to address this problem. Our approach synthesizes pseudo-ground truth from novel views that align with the limited training set and reintegrates these synthesized views into the training pipeline. Experimental results demonstrate that our bootstrapping technique not only reduces artifacts but also improves quantitative metrics. Furthermore, our technique is highly adaptable, allowing various Gaussian-based method to benefit from its integration. |
| title | Bootstrap-GS: Self-Supervised Augmentation for High-Fidelity Gaussian Splatting |
| topic | Graphics Artificial Intelligence Computer Vision and Pattern Recognition I.4.8 |
| url | https://arxiv.org/abs/2404.18669 |