A forensic facial examiner and professional team advantage for masked face identification

Face masks and coverings are often encountered by facial examiners in high stakes forensic case work. Forensic facial examiners (hereafter “facial examiners”) are trained to make face identifications and their decisions can be used as evidence in court. While research suggests that facial examiners can identify unconcealed faces with high accuracy, their identification performance for masked faces is unknown. Here we provide the first test of facial examiner performance for masked faces. An international sample of 61 facial examiners and 39 professional teams completed face identifications for 20 image pairs, which consisted of one unconcealed face image and one mask wearing face image. Examiners used their normal working procedures to make their identification decisions. The performance of facial examiners was compared against that of controls and six face identification algorithms. Both facial examiners and professional teams outperformed controls, however, professional teams made the least errors of all groups. The face identification algorithms achieved very high accuracy on the task. Our results support the use of facial examiners for the identification of masked faces and suggest a role for teams and human-machine working in applied practice. The findings back the notion that facial examiners use feature comparison, and this strategy is successful for matching pairs of images where one face wears a mask.