•  
  •  
 

Abstract

Project Mentor(s): Mary Radeke, PhD; Anthony Stahelski, PhD

Prior research on facial expression recognition (FER) has shown that the use of face masks significantly impairs the ability to correctly identify facial expressions, especially those that engage the lower regions of the face such as disgust and anger (Carbon, 2020; McCrackin et al. 2022). Not only does FER accuracy decrease with a masked or occluded face, but gaze patterns also shift and dwell time increases from those areas that are covered (mouth) to areas uncovered (eyes) (Rabadan, et al. 2022). In situations where both eyes and mouth are covered, the gaze patters shift and dwell time increases to the area between the eyes. Fixation patters and dwell times of occluded faces also differ depending on the facial expression. This study explored the effect of FER accuracy on eye scan patterns and fixation time. Using a repeated measures design, participants viewed images of anger, fear, sad, and neutral models with eyes and mouth occluded. Participants’ eye movements were recorded using Imotions eye tracking technology. Participants responded to an emotion recognition question to identify correct and incorrect assessments of the facial expressions. All facial expressions were divided into correct and incorrect recognition categories and collapsed across facial expressions. Eye scan patterns using Time to First Fixation measures (TTFF) and Dwell time (DT) were compared for the correct and incorrect categories.

SOURCE Form ID

154

Download

Share

COinS