Key points are not available for this paper at this time.
In this work, we present the world's first under-display lensless facial recognition system, which consists of a transparent micro-LED display, a specially designed mask for amplitude modulation, a CMOS sensor, and a deep learning model. By utilizing this kind of lensless optical component, the system can optically encrypt input facial information, ensuring that the light field information at the imaging plane is incomprehensible by humans. Compared to current technologies that encrypt the facial images, the advantage of this approach is that the system never captures any clear facial features, fundamentally protecting user privacy. To extract effective and generalizable features from these human-incomprehensible images, a recognition algorithm based on deep learning model is proposed. However, conventional deep learning models used for recognition systems have a fixed number of classes, necessitating retraining of the model during user registration or removal. To address this issue, we propose removing the output layer of the well-trained model and instead comparing the distance between each lensless image and the registered facial templates in the latent space for recognition tasks. This allows the system to successfully register and recognize new users without retraining the deep learning model. Our experimental results show that this system can provide stable recognition performance while preserving user privacy, with 93.02% accuracy, 97.51% precision, and 97.74% specificity.
Wu et al. (Sat,) studied this question.