Perspective Change The Face Recognition Under The Erp Study

Face recognition is an important function of human in society. Maybe we often do the recognition at anytime, from which we can get useful information. Face can provide abundant information then researches about face become the focus of social study. Previous studies of face recognition indicate that age, gender, race, expression, viewpoint and so on, they are all have the impact on face recognition, for the influence factors of the N170 which is the issue of argument about whether it is the specific element for face, such as Interstimulus interval, adapting time, different viewpoints are discussed deeply. In daily life, the face that we do the recognition is mostly for the horizontal direction, so, what's the face recognition mechanism under this situation? Whether N170 will appear or not? What are the differences between those situations? These are the contents that this study is to explore.This study includes two experiments:the first experiment is composed of a changed Oddball paradigm and a learning - recognition paradigm to test two problems:one is the inversion effect of face and whether the identity of face is being processed under the specific task; the second experiment we investigate the neural mechanism of face recognition across changes in viewpoint during adaptation paradigm. In this study we found the following results:1.From behavior data, it shows that in experiment 1, the accuracy and reaction time for face upside-down are higher and faster than for face upright. This result indicates that people always make a distinction according to the habituation, then, they will give negative response faster, because before giving the sure answer we need more details. And the EEG results show that, both the latency of P1 and N170 are longer for face upside-down to face upright, and the amplitude of P1 is bigger for face upside-down, but the amplitude of N170 shows reverse. These results show that inversion effect still exist under the situation of viewpoint changing, but face directions will affect the face recognition.2.The second part of the first experiment is recognition. Adding new faces into previous faces which are used to judge the orientation of faces, and the task is to judge whether the face is appeared in previous experiments and reaction with the keyboard, and a questionnaire will be done after experiment. Given the analysis about behavior data and EEG data, we think subject were not processing the face with consciousness in the first part of experiment 1. In the recognition task, the correct reject rate of new stimulus is higher than the other two types of faces, but the reaction time is longer; the accuracy of faces is the lowest and the reaction time is the longest for faces which views are inversed; the accuracy of new faces is the highest and the others' different are significantly. The ERP of hemispheres are different.3.In the adaption experiment, the amplitude of N170 in the right hemisphere is higher than the left, and the amplitude of N170 evoked by different faces is higher than the same. These results in favor of the early identification adaption effect on unfamiliar faces as in previous researches, and it shows the extraction of single face characterization occupied the early time of face processing. The amplitude of N170 evoked by frontal adaptive stimulus with deflecting testing stimulus is higher than that evoked by deflecting adaptive stimulus with frontal testing stimulus, which may indicate that more information is obtained after adapting to frontal faces, thus more brain areas are activated.The study presents the neural mechanisms of face recognition under the different viewpoints, and shows that both inversion effect and adaption effect exist. Compared with previous research results, the amplitude of both P1 and N170 delay for 30-50ms, it is probably for the reason of higher difficulty of the cognitive task; or it may also indicate that the direction process are included in the early stage of face process. In all, during face recognition, people will use different processing strategy based on different requirements, for example, holistic processing, feature processing, or even comprehensively use these two kinds.