Diffusion Bridges for 3D Point Cloud Denoising

Abstract

In this work, we address the task of point cloud denoising using a novel framework adapting Diffusion Schrödinger bridges to unstructured data like point sets. Unlike previous works that predict point-wise displacements from point features or learned noise distributions, our method learns an optimal transport plan between paired point clouds. In experiments on object datasets such as the PU-Net dataset and real-world datasets like ScanNet++ and ARKitScenes, improves by a notable margin over existing methods. Although our method demonstrates promising results utilizing solely point coordinates, we demonstrate that incorporating additional features like RGB information and point-wise DINOV2 features further improves the results.Code and pretrained networks are available at https://github.com/matvogel/P2P-Bridge.

Cite

Text

Hüni et al. "Diffusion Bridges for 3D Point Cloud Denoising." Proceedings of the European Conference on Computer Vision (ECCV), 2024. doi:10.1007/978-3-031-72627-9_11

Markdown

[Hüni et al. "Diffusion Bridges for 3D Point Cloud Denoising." Proceedings of the European Conference on Computer Vision (ECCV), 2024.](https://mlanthology.org/eccv/2024/huni2024eccv-diffusion/) doi:10.1007/978-3-031-72627-9_11

BibTeX

@inproceedings{huni2024eccv-diffusion,
  title     = {{Diffusion Bridges for 3D Point Cloud Denoising}},
  author    = {Hüni, Mathias Vogel and Tateno, Keisuke and Pollefeys, Marc and Tombari, Federico and Rakotosaona, Marie-Julie and Engelmann, Francis},
  booktitle = {Proceedings of the European Conference on Computer Vision (ECCV)},
  year      = {2024},
  doi       = {10.1007/978-3-031-72627-9_11},
  url       = {https://mlanthology.org/eccv/2024/huni2024eccv-diffusion/}
}