Activities per year
Abstract
In this work, we propose BP-MVSNet, a convolutional neural network (CNN)-based Multi-View-Stereo (MVS) method that uses a differentiable Conditional Random Field (CRF) layer for regularization. To this end, we propose to extend the BP layer [16] and add what is necessary to successfully use it in the MVS setting. We therefore show how we can calculate a normalization based on the expected 3D error, which we can then use to normalize the label jumps in the CRF. This is required to make the BP layer invariant to different scales in the MVS setting. In order to also enable fractional label jumps, we propose a differentiable interpolation step, which we embed into the computation of the pairwise term. These extensions allow us to integrate the BP layer into a multi-scale MVS network, where we continuously improve a rough initial estimate until we get high quality depth maps as a result. We evaluate the proposed BP-MVSNet in an ablation study and conduct extensive experiments on the DTU, Tanks and Temples and ETH3D data sets. The experiments show that we can significantly outperform the baseline and achieve state-of-the-art results.
Original language | English |
---|---|
Title of host publication | Proceedings - 2020 International Conference on 3D Vision, 3DV 2020 |
Publisher | IEEEXplore |
Pages | 394-403 |
Number of pages | 10 |
ISBN (Electronic) | 9781728181288 |
DOIs | |
Publication status | Published - 25 Nov 2020 |
Event | 8th International Conference on 3D Vision: 3DV 2020 - Online, Fukuoka, Virtual, Japan Duration: 25 Nov 2020 → 28 Nov 2020 http://3dv2020.dgcv.nii.ac.jp |
Publication series
Name | Proceedings - 2020 International Conference on 3D Vision, 3DV 2020 |
---|
Conference
Conference | 8th International Conference on 3D Vision |
---|---|
Abbreviated title | 3DV |
Country/Territory | Japan |
City | Fukuoka, Virtual |
Period | 25/11/20 → 28/11/20 |
Internet address |
Keywords
- CNN
- Conditional Random Field
- deep learning
- depth estimation
- Multi View Stereo
ASJC Scopus subject areas
- Artificial Intelligence
- Computer Vision and Pattern Recognition
Fingerprint
Dive into the research topics of 'BP-MVSNet: Belief-Propagation-Layers for Multi-View-Stereo'. Together they form a unique fingerprint.Activities
- 1 Poster presentation
-
Posterpräsentation International Conference on 3D Vision 2020
Sormann, C. (Speaker)
25 Nov 2020Activity: Talk or presentation › Poster presentation › Science to science