MatChA: Cross-Algorithm Matching with Feature Augmentation

TL;DR

This paper introduces MatChA, a novel method for cross-algorithm feature matching that enhances visual localization by augmenting and translating feature descriptors, addressing the challenge of different device-specific feature extraction algorithms.

Contribution

MatChA is the first approach to perform feature descriptor augmentation and translation for cross-detector feature matching, improving localization accuracy across different devices.

Findings

Significantly improves image matching in cross-feature scenarios

Enhances visual localization accuracy across benchmarks

Outperforms existing methods in cross-algorithm matching tasks

Abstract

State-of-the-art methods fail to solve visual localization in scenarios where different devices use different sparse feature extraction algorithms to obtain keypoints and their corresponding descriptors. Translating feature descriptors is enough to enable matching. However, performance is drastically reduced in cross-feature detector cases, because current solutions assume common keypoints. This means that the same detector has to be used, which is rarely the case in practice when different descriptors are used. The low repeatability of keypoints, in addition to non-discriminatory and non-distinctive descriptors, make the identification of true correspondences extremely challenging. We present the first method tackling this problem, which performs feature descriptor augmentation targeting cross-detector feature matching, and then feature translation to a latent space. We show that our method significantly improves image matching and visual localization in the cross-feature scenario and evaluate the proposed method on several benchmarks.