FreeSplatter: Pose-free Gaussian Splatting for Sparse-view 3D Reconstruction
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911130780172288 |
|---|---|
| author | Xu, Jiale Gao, Shenghua Shan, Ying |
| author_facet | Xu, Jiale Gao, Shenghua Shan, Ying |
| contents | Sparse-view reconstruction models typically require precise camera poses, yet obtaining these parameters from sparse-view images remains challenging. We introduce FreeSplatter, a scalable feed-forward framework that generates high-quality 3D Gaussians from uncalibrated sparse-view images while estimating camera parameters within seconds. Our approach employs a streamlined transformer architecture where self-attention blocks facilitate information exchange among multi-view image tokens, decoding them into pixel-aligned 3D Gaussian primitives within a unified reference frame. This representation enables both high-fidelity 3D modeling and efficient camera parameter estimation using off-the-shelf solvers. We develop two specialized variants--for object-centric and scene-level reconstruction--trained on comprehensive datasets. Remarkably, FreeSplatter outperforms several pose-dependent Large Reconstruction Models (LRMs) by a notable margin while achieving comparable or even better pose estimation accuracy compared to state-of-the-art pose-free reconstruction approach MASt3R in challenging benchmarks. Beyond technical benchmarks, FreeSplatter streamlines text/image-to-3D content creation pipelines, eliminating the complexity of camera pose management while delivering exceptional visual fidelity. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_09573 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | FreeSplatter: Pose-free Gaussian Splatting for Sparse-view 3D Reconstruction Xu, Jiale Gao, Shenghua Shan, Ying Computer Vision and Pattern Recognition Sparse-view reconstruction models typically require precise camera poses, yet obtaining these parameters from sparse-view images remains challenging. We introduce FreeSplatter, a scalable feed-forward framework that generates high-quality 3D Gaussians from uncalibrated sparse-view images while estimating camera parameters within seconds. Our approach employs a streamlined transformer architecture where self-attention blocks facilitate information exchange among multi-view image tokens, decoding them into pixel-aligned 3D Gaussian primitives within a unified reference frame. This representation enables both high-fidelity 3D modeling and efficient camera parameter estimation using off-the-shelf solvers. We develop two specialized variants--for object-centric and scene-level reconstruction--trained on comprehensive datasets. Remarkably, FreeSplatter outperforms several pose-dependent Large Reconstruction Models (LRMs) by a notable margin while achieving comparable or even better pose estimation accuracy compared to state-of-the-art pose-free reconstruction approach MASt3R in challenging benchmarks. Beyond technical benchmarks, FreeSplatter streamlines text/image-to-3D content creation pipelines, eliminating the complexity of camera pose management while delivering exceptional visual fidelity. |
| title | FreeSplatter: Pose-free Gaussian Splatting for Sparse-view 3D Reconstruction |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2412.09573 |