Perceived gaze direction and head orientation both impact the speed of face categorization, according to research results published in Vision Research.

In a recent experiment, researchers sought to determine how both perceived gaze direction and head orientation influence the human categorization of visual stimuli as faces. Currently, little is known about this influence. Researchers hypothesized that participants would respond faster to full-front faces and faces with a direct gaze compared with responses to left and right three-quarter faces and faces with an averted gaze, respectively. 

The experiment included 20 participants (14 women; mean age, 22.5±1.5 years), with results from 19 participants available for review. All participants had normal or corrected to normal visual acuity. 

Continue Reading

Color photographs of 137 faces and 618 non-face and non-face like objects across 112 categories were obtained, rescaled, and cropped to study parameters. Faces were included from multiple individuals across 7 poses, which included differing combinations of head orientation and perceived gaze direction. Participants were presented with a sequence of successive face and non-face images in a random order with an image duration of 91.7 milliseconds. Face images were both immediately preceded and followed by a random nonface object. 

Participants performed the experiment with high hit rates (95.4%±2.7% valid key presses in 100 to 1000 milliseconds from face onset) with few false alarms (2.2±3.5 random key presses). 

For each observer and each pose, response times for accurate key presses were averaged. Results of a 1-way repeated measures ANOVA compared average response times across all 7 poses, and results of post-hoc pairwise comparisons with Bonferroni correction demonstrated that average response time between full-front leftward and rightward gazes (mean difference, 9.91 milliseconds), three-quarter direct gaze leftward and rightward gazes (mean difference, 15.79 milliseconds), and three-quarter with averted gaze leftward and rightward gazes (mean difference, 0.89 milliseconds) were not significantly different. 

Investigators performed a 2-way repeated measures ANOVA to test whether head orientation and perceived gaze direction influenced response time in categorizing faces from other objects. Main effects for both were significant, although the interaction was not. Direct-gaze faces were, on average, categorized 48.36 milliseconds faster than averted-gaze faces, and full-front views were categorized 47.99 milliseconds faster than three-quarter side views.

A 1-way repeated measures ANOVA was performed to evaluate the effect of pose on participant minimum response time. The effect for pose was significant, with post-hoc pairwise comparisons indicating that the minimum response times for full-front direct gaze were “significantly faster” than response times for all three-quarter side view conditions. 

Although additional pairwise comparisons did not show significant differences, a trend showing minimum response times for three-quarter faces with averted gaze as the slowest among all poses was identified. 

In an effort to reduce outlier effects, researchers averaged the 3 earliest response times (6% of 50 face presentations) for each pose and observer. Results of a 1-way repeated measures ANOVA showed that the pose effect was significant, and post-hoc pairwise comparisons indicated that the averages of the 3 earliest response times were significantly faster than those for all three-quarter views. 

“The present study demonstrates that socially important cues of head orientation and perceived gaze direction produce comparable effects on real-time, behavioral face categorization during rapid serial visual presentation,” according to the research.


Or CC-F, Goh BK, Lee ALF. The roles of gaze and head orientation in face categorization during rapid serial visual presentation. Vision Res. 2021;188(11):65-73. doi:10.1016/j.visres.2021.05.012