Convolutional Hough Matching Networks

TL;DR

This paper introduces Convolutional Hough Matching, a neural network layer that leverages a Hough transform perspective for geometric matching, improving semantic visual correspondence under large intra-class variations.

Contribution

It presents a novel trainable neural layer based on Hough transform principles for geometric matching, enhancing robustness to image variations.

Findings

Sets new state-of-the-art on semantic visual correspondence benchmarks.

Demonstrates robustness to intra-class variations.

Effectively models non-rigid matching with few parameters.

Abstract

Despite advances in feature representation, leveraging geometric relations is crucial for establishing reliable visual correspondences under large variations of images. In this work we introduce a Hough transform perspective on convolutional matching and propose an effective geometric matching algorithm, dubbed Convolutional Hough Matching (CHM). The method distributes similarities of candidate matches over a geometric transformation space and evaluate them in a convolutional manner. We cast it into a trainable neural layer with a semi-isotropic high-dimensional kernel, which learns non-rigid matching with a small number of interpretable parameters. To validate the effect, we develop the neural network with CHM layers that perform convolutional matching in the space of translation and scaling. Our method sets a new state of the art on standard benchmarks for semantic visual correspondence, proving its strong robustness to challenging intra-class variations.