xEdgeFace: Efficient Cross-Spectral Face Recognition for Edge Devices

2025-04-28

AI Summary

To address cross-spectral face recognition (HFR) under resource-constrained edge-device scenarios—specifically, robust matching between thermal/near-infrared and visible-light images—this paper proposes a lightweight CNN-Transformer hybrid architecture. The method innovatively integrates local convolutional feature modeling with global cross-modal self-attention, augmented by an explicit cross-modal feature alignment mechanism, enabling end-to-end training with only a small number of paired heterogeneous samples. Compared to state-of-the-art methods, the model reduces computational complexity by 42% on average in FLOPs while maintaining competitive parameter efficiency. It achieves new state-of-the-art performance across multiple HFR benchmarks. Moreover, it preserves high accuracy on homogeneous RGB face recognition, demonstrating unified and efficient modeling for both heterogeneous and homogeneous face recognition tasks.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Learning on the Edge & Model Compression
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs

Abstract

Heterogeneous Face Recognition (HFR) addresses the challenge of matching face images across different sensing modalities, such as thermal to visible or near-infrared to visible, expanding the applicability of face recognition systems in real-world, unconstrained environments. While recent HFR methods have shown promising results, many rely on computation-intensive architectures, limiting their practicality for deployment on resource-constrained edge devices. In this work, we present a lightweight yet effective HFR framework by adapting a hybrid CNN-Transformer architecture originally designed for face recognition. Our approach enables efficient end-to-end training with minimal paired heterogeneous data while preserving strong performance on standard RGB face recognition tasks. This makes it a compelling solution for both homogeneous and heterogeneous scenarios. Extensive experiments across multiple challenging HFR and face recognition benchmarks demonstrate that our method consistently outperforms state-of-the-art approaches while maintaining a low computational overhead.

Problem

• Matching face images across different sensing modalities
• Reducing computational intensity for edge devices
• Enabling efficient training with minimal paired data

Innovation

• Hybrid CNN-Transformer architecture for HFR
• Efficient end-to-end training with minimal data
• Low computational overhead with strong performance