Purchase this article with an account.
Charles C.-F. Or, Talia Retter, Bruno Rossion; Automatic contribution of colour information to face categorization from briefly presented natural images. Journal of Vision 2016;16(12):1238. doi: 10.1167/16.12.1238.
Download citation file:
© 2017 Association for Research in Vision and Ophthalmology.
The contribution of colour to rapid categorization of natural images is debated. Here, the effect of colour on face categorization was examined using a recently validated paradigm (fast periodic visual stimulation) for measuring high-level face categorization responses with natural images (Rossion et al., 2015, J Vis). High-density electroencephalogram (EEG) was recorded during presentations of 50-s sequences of object images sinusoidally contrast-modulated at F = 12.0 Hz (i.e., 83-ms stimulus-onset asynchrony). Face images were embedded in the sequence at a fixed interval of F/9 (1.33 Hz). There were four conditions: 1) full-colour images; 2) greyscale images; 3) and 4) phase-scrambled images from Conditions 1 and 2, respectively, making faces and objects unrecognizable. Observers' task differed across two experiments: 20 observers responded to random colour changes of a fixation cross ("colour task"); another 20 observers responded when the fixation cross changed to a square ("shape task"). In both experiments, with natural images, selective responses to faces were recorded at 1.33 Hz and harmonics (2.67 Hz, etc.) over occipito-temporal areas, with right-hemisphere dominance; this response was absent with scrambled images. Importantly, in the shape task, face-categorization response (sum of all-channel-averaged responses at significant harmonics) was 22% stronger with natural images in colour than in greyscale (p = 0.025), indicating a substantial advantage from image colour information; this colour advantage was not significant in the colour task (p = 0.94). Behavioural analysis revealed that observers performing the colour task responded 20 ms slower when the natural images contained colour (p = 0.023), despite hit rates at ceiling (> 95% correct) in all conditions. However, no such response-time differences were found in the shape task (p = 0.71). Thus, the advantage of image colour to face categorization interacts with behaviour, suggesting that colour, when not a distractor, has an automatic contribution to face categorization in natural images.
Meeting abstract presented at VSS 2016
This PDF is available to Subscribers Only