Abstract
Visual image contains broadband information and is processed by different neural channels that are tuned to different spatial frequencies. Here we investigated whether or not our ability in gender identification on emotional faces was influenced by this early visual processing. Four types of emotional (happy, anger, sad, and fear) faces were used, and all of the stimuli were processed by spatial frequency analysis. Spatial frequency content in the original faces was filtered by using a high-pass filter (cut-off frequency was 24 cycles/image) for the HSF stimuli, and a low-pass filter (cut-off frequency was 6 cycles/image) for the LSF stimuli. Participants needed to identify the gender of the faces. The results showed that the participants responded faster and had higher accuracy to LSF faces than to HSF ones. They also responded faster and had higher accuracy to male faces than to female faces. Further analysis revealed that the identification of an angry man and a happy woman had advantage among combinations of genders and emotions in LSF conditions. However, this advantage was not manifested in HSF conditions. We concluded that the identification of gender with different emotions may rely on the processing of low spatial frequency channels.