Abstract
Object substitution masking (OSM) is an effective technique for limiting visual awareness of objects, including faces. Although OSM is thought to be engaged relatively late in visual representation, it is unclear whether it affects some higher-level visual processes more than others. Emotion perception is comprised of emotion detection (determining whether an emotion is present), and expression discrimination (categorizing that emotion). Because these processes are independent and depend on different facial information, it is possible that they may be differently impaired by OSM. We thus sought to clarify the timecourse of OSM for these two processes. Observers viewed four simultaneously presented faces, each of which was surrounded by four black dots and presented for 20-msec. Three of the faces were neutral, whereas the fourth face (the target) displayed an angry, happy, fearful or neutral expression. The location and the emotion of the target face were randomly selected on each trial. A red four-dot mask remained on the screen at the location of the target for 20, 60, 120, 240, or 640-msec after the faces disappeared. These masking dots served to disrupt perception of the target face, and to cue observers to report its expression. We separately evaluated the extent to which masks shown for increasingly long durations influenced the processes of emotion detection and expression discrimination. We found that OSM disrupted both emotion detection and expression discrimination. Notably, both processes were maximally disrupted with mask durations of 240-msec — roughly the same temporal profile observed in previous investigations of basic shape perception. However, unlike with shape perception, perception of emotion did not recover with longer mask durations, suggesting that once emotion is successfully masked by OSM, it is difficult to recover. Our results suggest that OSM is fully implemented at or prior to the beginning stages of emotion analysis.
Meeting abstract presented at VSS 2016