Non-frontal face recognition using GANs and memristor-based classifiers

2026-06-10

AI Summary

This work addresses the challenges of non-frontal face recognition on resource-constrained platforms—such as drones—where conventional deep learning approaches incur prohibitive computational costs. To this end, the authors propose a lightweight solution that uniquely integrates GAN-driven frontalization of facial poses with a memristor-based neuromorphic classifier. The proposed method achieves high recognition accuracy, reaching up to 96%, while substantially reducing both computational complexity and energy consumption. Its efficacy is validated on two standard benchmark datasets, demonstrating strong suitability for dynamic edge computing environments. By synergistically combining generative modeling with emerging neuromorphic hardware, this approach offers a novel and practical pathway toward efficient on-device face recognition under real-world constraints.

Abstract

Face recognition systems have advanced significantly through deep learning techniques, delivering high performance and robustness in complex scenarios. However, these approaches incur substantial computational overhead, limiting their in situ applicability in resource-constrained platforms such as drones, where they can address challenges including non-frontal facial imagery. Memristor-based neuromorphic systems have emerged as a compelling approach for edge AI applications, combining biologically inspired processing with efficient and scalable computation. In this work, we propose a facial recognition framework that addresses non-frontal pose variations by integrating lightweight generative adversarial network (GAN)-based pose frontalisation with memristor-based neuromorphic recognition. The experimental results on two datasets demonstrate the effectiveness of combining adversarial learning with memristive technology, achieving up to 96% identification accuracy. The proposed approach alleviates the computational bottlenecks of conventional AI and offers a scalable, efficient solution for face recognition in dynamic real-world environments.

Problem

• non-frontal face recognition
• resource-constrained platforms
• pose variation
• edge AI
• computational overhead

Innovation

• GAN-based pose frontalisation
• memristor-based neuromorphic computing
• non-frontal face recognition
• edge AI
• lightweight adversarial learning