The COVID pandemic raised concerns about our ability to recognize occluded faces because of impairments caused by face masks. Research has suggested that face occlusions impair our face recognition. Yin (1969) found that faces are harder to recognize upside-down. By showing a reduced inversion effect for masked faces, Freud et al. (2020) suggest that masking reduces holistic processing, although Fitousi et al. (2021) found mixed results for the inversion of occluded faces. Estudillo and Bindemann (2014) have suggested the type of task task affects recognition. The present study investigated the effect of inversion on faces occluded by masks and sunglasses and compared participants’ performance between memory and perception-based tasks across two experiments – face matching and recognition memory. Overall, results demonstrate that participants are better (faster and more accurate) at the matching task than the memory task. Furthermore, Experiment 2 demonstrated that occlusion reduced performance in both tasks, with little difference in performance between faces with masks and sunglasses. Interestingly, a reduced inversion effect (for accuracy and RT) was found for masked faces in the memory task; however, no such effect was found for the matching task. Hence, it is unclear whether reduced performance caused by occlusion is due to reduced holistic processing. Nevertheless, occlusion has different effects on the matching and the memory task. Based on the results, we suggest that in forensic settings, both forensic lineup and forensic face matching task should be used together.
Library of Congress Subject Headings
Face perception; Masks--Psychological aspects
Experimental Psychology (MS)
Department, Program, or Center
Pandurangan, Archana, "How Do We Recognize Masked Faces?" (2023). Thesis. Rochester Institute of Technology. Accessed from
RIT – Main Campus