Low‐resolution face recognition and the importance of proper alignment

Abstract

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.

Calibration of score based likelihood ratio estimation in automated forensic facial image comparison

“…Second, how to model the WSV and BSV distributions given the available data [16,39,45]? Given that the performance of facial image comparison highly depends on the quality of the data that a model is built with, the author in [34] suggests to use images having similar conditions to the real life facial image comparisons. Regarding the BSV modeling, [3] uses what is known as "pseudo-traces", that is using several pictures of the reference individual in the comparison instead of generic pictures of the same person not related to the case at hand.…”

Resolution Invariant Face Recognition Using a Distillation Approach

“…One possible solution is first to align the images in the HR domain and then down-sample to create the synthetic LR images. However, this approach would lead to optimistic results on synthetic LR [47] and it cannot be applied to native LR anyhow. Consequently, for low resolution facial landmark detection we train a CNN using Wing loss [48] on the WiderFace [49] dataset.…”

A novel generative adversarial network‐based super‐resolution approach for face recognition

“…In certain situations, such as ATM machines, facial recognition is utilized to detect and prevent fraudulent activities (Aru & Gozie, 2013). These examples highlight the diverse applications of facial recognition technology in enhancing security and access control across various domains (Long et al, 2022; Mikula et al, 2021; Peng et al, 2012; Sheng et al, 2023; Yang et al, 2022)…”

“…In certain situations, such as ATM machines, facial recognition is utilized to detect and prevent fraudulent activities (Aru & Gozie, 2013). These examples highlight the diverse applications of facial recognition technology in enhancing security and access control across various domains (Long et al, 2022;Mikula et al, 2021;Peng et al, 2012;Sheng et al, 2023;Yang et al, 2022) Images with high resolution are desirable and frequently necessary in most electronic imaging applications Qian et al, 2023;Sajjadi et al, 2017). A number of these applications use cameras and recording devices that produce low-resolution (LR) images and low-quality footage due to the devices being low-cost and using cheap equipment.…”