Purchase this article with an account.
Haojiang Ying, Wenxuan Cheng, Hong Xu; Ensemble Coding of Facial Attractiveness is Largely Driven by the High Spatial Frequency Information. Journal of Vision 2019;19(10):196. doi: https://doi.org/10.1167/19.10.196.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Perception of facial attractiveness is shaped by our visual experience and by context. Researchers recently showed that our visual system is capable of perceiving the facial attractiveness of a group of faces via ensemble coding. However, the mechanism of ensemble coding of facial attractiveness remains largely unknown. Specifically, does our visual system summaries the high spatial frequency information from the face, or the low spatial frequency information? To answer this question we used adaptation paradigm to scrutinize the spatial frequency in ensemble representation of facial attractiveness. Participants were asked to judge the attractiveness of the test faces, after adapting to four unattractive faces. In each block, participants were exposed to different kind of faces:1) four unattractive faces (full bandwidth; FB), 2) the high spatial frequency version (HSF; > 32 cycles per face), and 3) the low spatial frequency version (LSF; < 8 cycles per face) of them. Results suggested that compared to the non-adaptation baseline, both FB condition (M = 7.58%, SEM = 1.77%; t(29) = 4.28, p < .001, Cohen’s d = 1.59) and the HSF condition (M = 6.05%, SEM = 21.93%; t(29) = 3.13, p = .004, Cohen’s d = 1.16) generated significant and similar adaptation aftereffects; while the LSF condition (M = 1.87%, SEM = 1.32%; t(29) = 1.42, p = .166, Cohen’s d = 0.53) failed to yield a significant aftereffect. Our results suggest that the ensemble coding of facial attractiveness is largely driven by the high spatial frequency information from the faces. This finding opens the question for the neural and computational mechanisms of ensemble coding.
This PDF is available to Subscribers Only