The Effects Of Semantic Information And Name On The Face Processing

The study of face identification has been a hot topic in the field of face recognition.Bruce and Young proposed the functional modesl of face recognition, In the model there are three,distinct sequential stages involved in identifying an individual from his or her face, following thestructure coding, perception characterization activation, the extraction of semanticinformation and the extraction of the name. However Burton et al proposed the interactiveactivation model.The model suggested that facaial identity recognition is not just a serialprocess, but there are mutual promotion between each processing stage.The model alsosuggested that the name code occurred in the early stage of the face recognition, not at thelast.The two models have different opinions in the extration of face recognition.The hightemporal resolution of ERP provide technical support for the process of information extractionin different capacities of the face identification.ERP technique was used in the study. In theexperiment,participants were asked to learn faces with different identity information, furtherinvestigating the differences on ERPs components of the subjects’ reaction with differentfaces identity information, in order to verify the model therory.The study used the “Learning--Test”paradigm to investigate ERP differences in theextraction of the semantic information and name in the face recognition process through twoexperiments. In experiment1, subjects were asked to learn the information of face containingcareer information or not containing career information, in the subsequent test phase, thesubjects were allowed to react to different types of face stimuli, thereby we can find thedifferences of behavioral data and the ERPs components caused by occupation information. Inexperiment2, subjects were asked to learn the faces which have names and faces which haveno name, thereby we can find the differences of behavioral data and ERPs components causedby name.The experiments were the single factor within subjects experimental design.The results showed that:(1) In the behavioral data, learned face with occupationinformation was faster than learned face without occupation information in the reaction time;learned face with name information was faster than learned face without name information in the reaction time. The study semantic information and name promoted the recogniton offace.(2) In the ERP data, the difference is remarkable in the N400component between theface containing career information and the face not containing career information. N400component reflected the extration of semantic information.(3) The face which has names hasremarkable differences than face which has no name in the N250component. The effect ofN250in the temporo-occipitalconfirmed name has an effect on the facial familiarty. Nameinformation promoted the recognition of face.In the behavioral studies, the semantic information and name promoted the recognitionof face, which is consistent with the view of the interactive activation model which suggeststhere are the promoting effect in the different facial processing units. ERP results showed theeffect of name is earlier than semantic information in the recognition of face in the course oftime, which also is consistent with the view of the interactive activation model. The data ofbehavior and ERP are consisted with the view of the interactive activation model.