The Processing Of Face Identity And Facial Expression:an ERP Approach

Face recognition is a new research field and attracts more and more attention. Face recognition is different with other objects and is a relatively independent process. This paper puts forward a new approach on face images processing to achieve a unified standard. Then we carry out the face recognition and facial expression recognition experiments and experimental study of familiar faces.First, this paper introduces how to deal with the face pictures. This paper puts forward a new way of processing face images as a group:Through the basic image processing, the equalization and edge gradient image processing, all face images reach the same effect. If you need to add a new image, you need to select anyone in the face image database as reference to deal based on the matching process. The processing procedure of the face images does not influence the characteristics of the image itself as far as possible to eliminate the interference factors.Then we use the ERP (Event-related potentials) technology to carry out a face recognition experiment. Face recognition is divided into face identity recognition and facial expression recognition. There a dispute between the independent and the contact of the two, but studies have shown that there are some relations between the two. In this paper, using four-quadrant paradigm, we research the relations and differences between face identity and facial expression recognition. In this experiment, we use the fear expression and neutral expression of unfamiliar faces of men and women. Behavioral results show that:the accuracy of face identity recognition is higher than the facial expression recognition, especially to identify this "afraid" negative expression, but the reaction time is not significantly different. EEG results show that:In the early ERP, face identity recognition and facial expression recognition is no significant difference, but in the late, in the face identity recognition we discovery the wave separation of fear expression and neutral expression. After300ms, we find the wave separation on the left attention and right attention, indicating the face identity recognition affects attention. But facial expression recognition does not appear the influence above. We see that face recognition is easier than the expression recognition, the recognition of fearful expression is easier than neutral facial expression, face recognition will be affected by the expression, the recognition of face and expression have a certain contact, and face recognition appears skewed popularity.Finally, we carry out the experimental study of familiar faces, familiar degree is an important factor affecting the face recognition. We has done a lot of research and experiments in the conscious perception. This paper use the CFS (continuous flash suppression) paradigm to research the strange, familiar famous and our own face in the conscious suppressing. The behavioral data is mainly used to judge whether the participants really see or not see, as the reference data. The EEG results show that:face images in conscious suppressing can also cause brain electric wave, but the brain wave of face images in the conscious perception is larger than the wave in the conscious suppressing. We can also see that in the conscious perception the wave of ERP evoked by familiar faces is smaller than other level of familiarity faces. We find that in the conscious suppressing the wave is also affected by the faces of different familiarity, and there are fluctuations in the brain. But because the intensity of the stimulus is relatively weak, so there is not so obvious as the wave in the conscious perception.