IM-3D: Iterative Multiview Diffusion and Reconstruction for High-Quality 3D Generation
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866913233709826048 |
|---|---|
| author | Melas-Kyriazi, Luke Laina, Iro Rupprecht, Christian Neverova, Natalia Vedaldi, Andrea Gafni, Oran Kokkinos, Filippos |
| author_facet | Melas-Kyriazi, Luke Laina, Iro Rupprecht, Christian Neverova, Natalia Vedaldi, Andrea Gafni, Oran Kokkinos, Filippos |
| contents | Most text-to-3D generators build upon off-the-shelf text-to-image models trained on billions of images. They use variants of Score Distillation Sampling (SDS), which is slow, somewhat unstable, and prone to artifacts. A mitigation is to fine-tune the 2D generator to be multi-view aware, which can help distillation or can be combined with reconstruction networks to output 3D objects directly. In this paper, we further explore the design space of text-to-3D models. We significantly improve multi-view generation by considering video instead of image generators. Combined with a 3D reconstruction algorithm which, by using Gaussian splatting, can optimize a robust image-based loss, we directly produce high-quality 3D outputs from the generated views. Our new method, IM-3D, reduces the number of evaluations of the 2D generator network 10-100x, resulting in a much more efficient pipeline, better quality, fewer geometric inconsistencies, and higher yield of usable 3D assets. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_08682 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | IM-3D: Iterative Multiview Diffusion and Reconstruction for High-Quality 3D Generation Melas-Kyriazi, Luke Laina, Iro Rupprecht, Christian Neverova, Natalia Vedaldi, Andrea Gafni, Oran Kokkinos, Filippos Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning Most text-to-3D generators build upon off-the-shelf text-to-image models trained on billions of images. They use variants of Score Distillation Sampling (SDS), which is slow, somewhat unstable, and prone to artifacts. A mitigation is to fine-tune the 2D generator to be multi-view aware, which can help distillation or can be combined with reconstruction networks to output 3D objects directly. In this paper, we further explore the design space of text-to-3D models. We significantly improve multi-view generation by considering video instead of image generators. Combined with a 3D reconstruction algorithm which, by using Gaussian splatting, can optimize a robust image-based loss, we directly produce high-quality 3D outputs from the generated views. Our new method, IM-3D, reduces the number of evaluations of the 2D generator network 10-100x, resulting in a much more efficient pipeline, better quality, fewer geometric inconsistencies, and higher yield of usable 3D assets. |
| title | IM-3D: Iterative Multiview Diffusion and Reconstruction for High-Quality 3D Generation |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2402.08682 |