Time course and neural locus of the Flashed Face Distortion Effect

Abstract

Viewing a rapid sequence of face images shown in the periphery can lead to large caricature-like distortions in the perceived images, a phenomenon known as the Flashed Face Distortion Effect (FFDE). The mechanisms underlying FFDE are poorly understood. Here we examined the timing and sites of the adaptation processes giving rise to the FFDE. To investigate the effects of presentation rate, we maintained consistent trial lengths while assessing how variations in the temporal frequencies of face presentation influenced the magnitude of face distortion and the averaging of facial expressions. Over a wide range of temporal frequencies (1.2-60 Hz) tested, we observed a decrease in FFDE strength as the presentation rate increased. To probe the neural sites of FFDE, we varied whether successive faces were presented to the same or different eyes using a dichoptic display. Distortion effects were comparable for monocular, binocular, and interocular conditions, yet much larger than a control condition where faces were presented with a temporal interval between successive images, suggesting a cortical locus for FFDE.

Keywords: Averaging; Face distortion; Interocular transfer; Time course.

Figure 1

Example trial for the distortion rating task and the facial expression judgment task. The left panel illustrates an example trial of the distortion rating task. Neutral faces were used. Participants were asked to estimate overall how distorted they thought the faces were. The right panel shows an example trial of the facial expression judgment task. In the example trial shown, the average facial expression was 50 in the left visual field and 35 in the right visual field. The participants were asked to determine which visual field, left or right, exhibited a happier average facial expression. For both the facial expression judgment task and the distortion rating task, the temporal rates of face presentation were varied between 1.2, 4.8, 24, and 60 Hz across trials.

Figure 2

The strength of FFDE and performance of facial expression averaging as a function of face presentation rate. The strength of the FFDE decreased with the face presentation rate (left panel) whereas the averaging performance increased with the face presentation rate. Error bars indicate SEM.

Figure 3

Illustrations of conditions in Experiment 2. In the monocular condition, faces were presented to either the left or right eye. In the binocular condition, faces were simultaneously presented to both eyes. In the interocular condition, successive faces were presented alternately to different eyes. In the control condition, successive faces were presented to either the left or right eye but interspersed with a gray field of equivalent duration between each face presentation. Note that the figure illustrates presentation of successive face images; the actual number of face images were more extended, as detailed in the Methods section.

Figure 4

Mean and individual distortion ratings for the four viewing conditions in Experiment 2. The distortion ratings were comparable among the monocular, binocular, and interocular conditions, but all were significantly higher than those in the control condition. Asterisks represent the averaged results and empty circles indicate individual data. Error bars represented SEM.