Brain Research. Cognitive Brain Research

Publication date: 2003-05
Volume: 16 Pages: 416 - 424
ISSN: 0926-6410, 1872-6348 PMID: 12706221
DOI: 10.1016/S0926-6410(03)00056-9
Publisher: Elsevier Science Publishers


Goffaux, Valérie
Gauthier, Isabel ; Rossion, Bruno


Adult, Electroencephalography, Electrophysiology, Evoked Potentials, Face, Female, Form Perception, Humans, Male, Visual Perception, Science & Technology, Technology, Life Sciences & Biomedicine, Computer Science, Artificial Intelligence, Neurosciences, Neuroimaging, Computer Science, Neurosciences & Neurology, face processing, specificity, N170, configuration, spatial scales, EVENT-RELATED POTENTIALS, INVERTED FACES, CONFIGURAL INFORMATION, DOMAIN SPECIFICITY, RECOGNITION, EXPERTISE, INVERSION, PERCEPTION, CATEGORIZATION, MECHANISMS, Experimental Psychology


Event-related potential (ERP) studies have highlighted an occipito-temporal potential, the N170, which is larger for faces than for other categories and delayed by stimulus inversion of faces, but not of other objects. We examined how high-pass and low-pass filtering modulate such early differences between the processing of faces and objects. Sixteen grey-scale pictures of faces and cars were filtered to preserve only relatively low (LSF) or high (HSF) spatial frequencies and were presented upright or upside-down. Subjects reported the orientation of the faces and cars in broad-pass and filtered conditions. In the broad-pass condition, we replicated typical N170 face-specific effects of amplitude and delay with inversion. These effects were also present in the LSF condition. However, a completely different pattern was revealed by the HSF condition: (1). a similar N170 amplitude for cars as compared to faces and (2). an absence of N170 latency delay with face inversion. These results show that the source of early processing differences between faces and objects is related to the extraction of information contained mostly in the LSF, which conveys coarse configuration cues particularly salient for face processing.