Lightweight Cross-Spectral Face Recognition via Contrastive Alignment and Distillation

TL;DR

This paper presents a lightweight hybrid CNN-Transformer framework for heterogeneous face recognition that is efficient, adaptable to resource-limited devices, and achieves competitive performance with minimal paired data.

Contribution

It adapts a hybrid CNN-Transformer model for HFR, enabling efficient training with limited data and maintaining high accuracy on benchmarks.

Findings

Achieves state-of-the-art or competitive results on HFR benchmarks.

Maintains low computational requirements suitable for edge devices.

Effective with limited paired heterogeneous face data.

Abstract

Heterogeneous Face Recognition (HFR) aims at matching face images captured across different sensing modalities, such as thermal-to-visible or near-infrared-to-visible, enhancing the usability of face recognition systems in challenging real-world conditions. Although recent HFR methods have achieved significant improvements in performance, many rely on computationally expensive models, making them impractical for deployment on resource-limited edge devices. In this work, we introduce a lightweight yet effective HFR framework by adapting a hybrid CNN-Transformer model originally developed for RGB homogeneous face recognition. Our approach enables efficient end-to-end training with only a small amount of paired heterogeneous data, while still maintaining strong performance on standard RGB face recognition benchmarks. This makes it suitable for both homogeneous and heterogeneous settings. Comprehensive experiments on several challenging HFR and face recognition benchmarks show that our method achieves state-of-the-art or competitive performance while keeping computational requirements low.