88‐4: <i>Distinguished Student Paper:</i> Enhancing Face Recognition Accuracy for Under‐Display Cameras via Image Restoration

Under‐display Camera (UDC) allows for a full‐screen display by placing the camera beneath the screen, eliminating the need for display holes. However, diffraction caused by display pixels results in the degradation of UDC image quality, including reduced transmittance, noise, blur, and flare. This leads to decreased facial recognition accuracy, compromising UDC's practicality. While recent research has focused on image restoration for UDC, studies on UDC facial recognition remain insufficient. Previous UDC face recognition studies use synthetic UDC datasets without flare to train degradation models and apply them to high‐quality datasets of human faces. However, they do not address real UDC datasets or Face ID performance, a common smartphone application. This paper presents three UDC facial recognition datasets generated by a generative adversarial network (GAN) trained on synthetic and real‐world datasets with flares. We aim to analyze the impact of UDC restoration on Face ID accuracy, contributing to the expansion of UDC face recognition research.