Convolutional Neural Network Architecture for Geometric Matching

Ignacio Rocco, Relja Arandjelovic, Josef Sivic; The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, pp. 6148-6157

Abstract


We address the problem of determining correspondences between two images in agreement with a geometric model such as an affine or thin-plate spline transformation, and estimating its parameters. The contributions of this work are three-fold. First, we propose a convolutional neural network architecture for geometric matching. The architecture is based on three main components that mimic the standard steps of feature extraction, matching and simultaneous inlier detection and model parameter estimation, while being trainable end-to-end. Second, we demonstrate that the network parameters can be trained from synthetically generated imagery without the need for manual annotation and that our matching layer significantly increases generalization capabilities to never seen before images. Finally, we show that the same model can perform both instance-level and category-level matching giving state-of-the-art results on the challenging Proposal Flow dataset.

Related Material


[pdf] [arXiv] [poster] [video]
[bibtex]
@InProceedings{Rocco_2017_CVPR,
author = {Rocco, Ignacio and Arandjelovic, Relja and Sivic, Josef},
title = {Convolutional Neural Network Architecture for Geometric Matching},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {July},
year = {2017}
}