Purchase this article with an account.
Daniel Fiset, Josiane Leclerc, Jessica Royer, Valérie Plouffe, Caroline Blais; Facial expressions modulate visual features utilization in unfamiliar face identification. Journal of Vision 2016;16(12):161. doi: https://doi.org/10.1167/16.12.161.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Flexible face identification requires extracting expression-independent visual information while leaving aside expression-dependant visual features. Most studies in the field suggest that the eye area is by far the most important visual feature for face identification (e.g. Butler et al., 2010; Schyns, Bonnar & Gosselin, 2002; Sekuler, Gaspar, Gold & Bennett, 2004). However, these studies were mostly done with stimuli showing only one (e.g. neutral) or two (e.g. neutral and happy) facial expressions. Here, we investigated the impact of facial expressions on the utilization of facial features in a facial identification task with unfamiliar faces (10 identities; 5 female). Each identity showed six different facial expressions (anger, disgust, fear, happy, neutral, sad). We used the Bubbles technique (Gosselin & Schyns, 2001) to reveal the diagnostic visual features in five non-overlapping spatial frequency bands. Twenty-five participants first learned to recognize the identities until their performance reached 95% correct for each facial expression. After reaching this performance criterion, they performed 1320 trials (220 for each facial expression) with bubblized stimuli. For each facial expression, the number of bubbles was adjusted on a trial-by-trial basis to maintain a correct identification rate of 55%. Overall, we closely replicate other studies (Caldara et al., 2005; Schyns, Bonnar & Gosselin, 2002). However, when each facial expression was analysed independently, the results show clear inter-expression differences. For neutral faces, the eyes are by far the most diagnostic features. However, for other facial expressions, participants showed a clear processing bias for expression-dependant facial features (e.g. the mouth for happy and disgust). In short, our data suggests that facial expression features are closely bound to identification in unfamiliar face recognition.
Meeting abstract presented at VSS 2016
This PDF is available to Subscribers Only