Abstract
An event related potential (ERP) study was conducted to evaluate the effects of Gabor filtered face images on the N170 Event Related Potential (ERP) component. Gabor filtered images are of interest for two reasons. First, Gabor filtering has proven to be an effective first step in several face recognition algorithms developed by the computer vision community; it is therefore interesting to see how the human vision system responds to Gabor-filtered images. Second, previous research has indicated that face images filtered with low band pass filters are preferred to face images filtered with high pass filters. We were interested in the effect these filtered images had on the well define N170 component that is associated with face recognition with “normal” face stimuli in humans. Comparisons were made to achromatic “normal” photographs of faces, in a rapid presentation task. ERP' were recorded using a high density net in a passive viewing task. Participants viewed random presentations of either “normal” faces or Gabor filtered faces. Our results showed that for the “Gabor face” condition N170 was delayed in latency, and was reduced in amplitude, this difference was statistically significant. This is contrary to previous research indicating that low band pass filters in face recognition yield better speed and accuracy performance compared to high bands pass filters.
National Science Foundation Grant, The National Defense Agency, & Colorado State University - Research Experience for Undergraduates In Mind Brain and Behavior.