Search

Face recognition methods for low resolution are often developed and tested on down‐sampled images instead of on real low‐resolution images. Although there is a growing awareness that down‐sampled and real low‐resolution images are different, few efforts have been made to analyse the differences in recognition performance. Here, the authors explore the differences and demonstrate that alignment is a major cause, especially in the absence of pose and illumination variations. The authors found that the recognition performances on down‐sampled images are flattered mostly due to the fact that the images are perfectly aligned before down‐sampling using high‐resolution landmarks, while the real low‐resolution images have much poorer alignment. To obtain better alignment for real low‐resolution images, the authors apply matching score‐based registration which does not rely on accurate landmarks. The authors propose to divide low resolution into three ranges to harmonise the terminology: upper low resolution (ULR), moderately low resolution (MLR), and very low resolution (VLR). Most face recognition methods perform well on ULR. MLR is a challenge for commercial systems, but a low‐resolution deep‐learning method can handle it very well. The performance of most methods degrades significantly for VLR, except for simple holistic methods which perform the best. [ABSTRACT FROM AUTHOR]

A very common case for law enforcement is recognition of suspects from a long distance or in a crowd. This is an important application for low‐resolution face recognition (in the authors' case, face region below 40 × 40 pixels in size). Normally, high‐resolution images of the suspects are used as references, which will lead to a resolution mismatch of the target and reference images since the target images are usually taken at a long distance and are of low resolution. Most existing methods that are designed to match high‐resolution images cannot handle low‐resolution probes well. In this study, they propose a novel method especially designed to compare low‐resolution images with high‐resolution ones, which is based on the log‐likelihood ratio (LLR). In addition, they demonstrate the difference in recognition performance between real low‐resolution images and images down‐sampled from high‐resolution ones. Misalignment is one of the most important issues in low‐resolution face recognition. Two approaches – matching‐score‐based registration and extended training of images with various alignments – are introduced to handle the alignment problem. Their experiments on real low‐resolution face databases show that their methods outperform the state‐of‐the‐art. [ABSTRACT FROM AUTHOR]