Learned Multi-Patch Similarity

Abstract

Estimating a depth map from multiple views of a scene is a fundamental task in computer vision. As soon as more than two viewpoints are available, one faces the very basic question how to measure similarity across >2 image patches. Surprisingly, no direct solution exists, instead it is common to fall back to more or less robust averaging of two-view similarities. Encouraged by the success of machine learning, and in particular convolutional neural networks, we propose to learn a matching function which directly maps multiple image patches to a scalar similarity score. Experiments on several multi-view datasets demonstrate that this approach has advantages over methods based on pairwise patch similarity.

Cite

Text

Hartmann et al. "Learned Multi-Patch Similarity." International Conference on Computer Vision, 2017. doi:10.1109/ICCV.2017.176

Markdown

[Hartmann et al. "Learned Multi-Patch Similarity." International Conference on Computer Vision, 2017.](https://mlanthology.org/iccv/2017/hartmann2017iccv-learned/) doi:10.1109/ICCV.2017.176

BibTeX

@inproceedings{hartmann2017iccv-learned,
  title     = {{Learned Multi-Patch Similarity}},
  author    = {Hartmann, Wilfried and Galliani, Silvano and Havlena, Michal and Van Gool, Luc and Schindler, Konrad},
  booktitle = {International Conference on Computer Vision},
  year      = {2017},
  doi       = {10.1109/ICCV.2017.176},
  url       = {https://mlanthology.org/iccv/2017/hartmann2017iccv-learned/}
}