Non-frontal face recognition using GANs and memristor-based classifiers
Researchers have developed a novel face recognition system that combines Generative Adversarial Networks (GANs) with memristor-based classifiers to handle non-frontal facial images. This approach aims to reduce the computational overhead of traditional AI methods, making them suitable for edge devices like drones. The system achieved up to 96% identification accuracy on two datasets by using GANs for pose frontalization and memristors for efficient recognition. AI
IMPACT This research could enable more efficient and accurate face recognition on edge devices, expanding applications in areas like autonomous systems and surveillance.