Learning to Compare Image Patches via Convolutional Neural Networks

Sergey Zagoruyko (1, 2, 3) , Nikos Komodakis (2, 1, 3, 4)

1 LIGM - Laboratoire d'Informatique Gaspard-Monge
2 ENPC - École des Ponts ParisTech
3 imagine [Marne-la-Vallée]
4 CSD-UOC - Computer Science Department [Crete]

In this paper we show how to learn directly from image data (i.e., without resorting to manually-designed features) a general similarity function for comparing image patches, which is a task of fundamental importance for many computer vision problems. To encode such a function, we opt for a CNN-based model that is trained to account for a wide variety of changes in image appearance. To that end, we explore and study multiple neural network architectures, which are specifically adapted to this task. We show that such an approach can significantly outperform the state-of-the-art on several problems and benchmark datasets.

Mots clés: Computational modeling; Convolutional codes; Benchmark testing; Adaptation models; Neural networks; Training; Computer architecture

Domaines: Vision par ordinateur et reconnaissance de formes [cs.CV]; Réseau de neurones [cs.NE]

Sergey Zagoruyko, Nikos Komodakis. Learning to Compare Image Patches via Convolutional Neural Networks. IEEE Conference on Computer Vision and Pattern Recognition 2015, Jun 2015, Boston, United States. ⟨10.1109/CVPR.2015.7299064⟩. ⟨hal-01246261⟩