From Modalities to Styles: Rethinking the Domain Gap in Heterogeneous Face Recognition

TL;DR

This paper introduces a novel feature modulation technique called CAIM that adapts face recognition models to different modalities, significantly improving heterogeneous face recognition performance with limited data.

Contribution

The paper proposes the CAIM module, a new style modulation approach that enhances existing face recognition networks for heterogeneous face recognition tasks.

Findings

Outperforms state-of-the-art HFR methods on multiple benchmarks

Enables end-to-end training with limited paired samples

Effectively bridges domain gaps between different face modalities

Abstract

Heterogeneous Face Recognition (HFR) focuses on matching faces from different domains, for instance, thermal to visible images, making Face Recognition (FR) systems more versatile for challenging scenarios. However, the domain gap between these domains and the limited large-scale datasets in the target HFR modalities make it challenging to develop robust HFR models from scratch. In our work, we view different modalities as distinct styles and propose a method to modulate feature maps of the target modality to address the domain gap. We present a new Conditional Adaptive Instance Modulation (CAIM ) module that seamlessly fits into existing FR networks, turning them into HFR-ready systems. The CAIM block modulates intermediate feature maps, efficiently adapting to the style of the source modality and bridging the domain gap. Our method enables end-to-end training using a small set of paired samples. We extensively evaluate the proposed approach on various challenging HFR benchmarks, showing that it outperforms state-of-the-art methods. The source code and protocols for reproducing the findings will be made publicly available