作者: Kirsten Petras , Sanne ten Oever , Christianne Jacobs , Valerie Goffaux
DOI: 10.1016/J.NEUROIMAGE.2018.10.086
关键词: Electroencephalography 、 Information integration 、 Stimulus (physiology) 、 Face (geometry) 、 Artificial intelligence 、 High spatial frequency 、 Coarse to fine 、 Perception 、 Pattern recognition 、 Computer science
摘要: Abstract Coarse-to-fine theories of vision propose that the coarse information carried by low spatial frequencies (LSF) visual input guides integration finer, high frequency (HSF) detail. Whether and how LSF modulates HSF processing in naturalistic broad-band stimuli is still unclear. Here we used multivariate decoding EEG signals to separate respective contribution neural response evoked images. Participants viewed images human faces, monkey faces phase-scrambled versions were either or filtered contain HSF. We trained classifiers on scalp-patterns scrambled evaluated derived models intact trials. found reduced when was informative towards image content, indicating does guide fine detail, line with coarse-to-fine theories. discuss potential cortical mechanisms underlying such feedback.