BP-MVSNet: Belief-Propagation-Layers for Multi-View-Stereo

Christian Sormann*, Patrick Knöbelreiter, Andreas Kuhn, Mattia Rossi, Thomas Pock, Friedrich Fraundorfer

*Korrespondierende/r Autor/-in für diese Arbeit

Publikation: Beitrag in Buch/Bericht/KonferenzbandBeitrag in einem KonferenzbandBegutachtung


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.

TitelProceedings - 2020 International Conference on 3D Vision, 3DV 2020
Herausgeber (Verlag)IEEEXplore
ISBN (elektronisch)9781728181288
PublikationsstatusVeröffentlicht - 25 Nov. 2020
VeranstaltungInternational Virtual Conference on 3D Vision: 3DV 2020 - Virtual, Fukuoka, Japan
Dauer: 25 Nov. 202028 Nov. 2020


NameProceedings - 2020 International Conference on 3D Vision, 3DV 2020


KonferenzInternational Virtual Conference on 3D Vision
OrtVirtual, Fukuoka

ASJC Scopus subject areas

  • Artificial intelligence
  • Maschinelles Sehen und Mustererkennung


Untersuchen Sie die Forschungsthemen von „BP-MVSNet: Belief-Propagation-Layers for Multi-View-Stereo“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren