PN-Net: Conjoined Triple Deep Network for Learning Local Image Descriptors

TL;DR

This paper introduces PN-Net, a CNN-based local image descriptor that achieves high matching accuracy with low computational cost by using triplet training and a novel loss function, outperforming existing methods in speed and memory efficiency.

Contribution

PN-Net presents a new CNN-based local descriptor trained with triplets and a novel loss function, offering improved performance and efficiency over prior descriptors.

Findings

Outperforms MatchNet and DeepCompare in accuracy.

Extraction time comparable to binary descriptors like BRIEF and ORB.

Significantly reduced training and execution time.

Abstract

In this paper we propose a new approach for learning local descriptors for matching image patches. It has recently been demonstrated that descriptors based on convolutional neural networks (CNN) can significantly improve the matching performance. Unfortunately their computational complexity is prohibitive for any practical application. We address this problem and propose a CNN based descriptor with improved matching performance, significantly reduced training and execution time, as well as low dimensionality. We propose to train the network with triplets of patches that include a positive and negative pairs. To that end we introduce a new loss function that exploits the relations within the triplets. We compare our approach to recently introduced MatchNet and DeepCompare and demonstrate the advantages of our descriptor in terms of performance, memory footprint and speed i.e. when run in GPU, the extraction time of our 128 dimensional feature is comparable to the fastest available binary descriptors such as BRIEF and ORB.