Learning-based Natural Geometric Matching with Homography Prior

TL;DR

This paper introduces a learning-based geometric matching method that emphasizes both alignment accuracy and naturalness, utilizing homography prior, Pearson correlation, and a novel loss function, outperforming existing methods.

Contribution

The authors propose a new geometric matching approach combining homography prior with a novel loss function, enhancing both naturalness and accuracy in image alignment.

Findings

Outperforms state-of-the-art methods in alignment accuracy.

Improves naturalness of transformed images.

Effective handling of large viewpoint variations.

Abstract

Geometric matching is a key step in computer vision tasks. Previous learning-based methods for geometric matching concentrate more on improving alignment quality, while we argue the importance of naturalness issue simultaneously. To deal with this, firstly, Pearson correlation is applied to handle large intra-class variations of features in feature matching stage. Then, we parametrize homography transformation with 9 parameters in full connected layer of our network, to better characterize large viewpoint variations compared with affine transformation. Furthermore, a novel loss function with Gaussian weights guarantees the model accuracy and efficiency in training procedure. Finally, we provide two choices for different purposes in geometric matching. When compositing homography with affine transformation, the alignment accuracy improves and all lines are preserved, which results in a more natural transformed image. When compositing homography with non-rigid thin-plate-spline transformation, the alignment accuracy further improves. Experimental results on Proposal Flow dataset show that our method outperforms state-of-the-art methods, both in terms of alignment accuracy and naturalness.