Abstract: Human faces convey important messages, such as identity, age, sex, and emotional expression. The traditional view of facial expressions is that certain configurations of facial muscle contractions convey information about the emotional state of a person. Certain facial expressions and certain emotions are intimately tied together. There are many studies emphasizing on the influencing factor of facial expression recognition. The results indicate that influencing factors include age (Pfütze, 2002), emotion condition (Mullins, 2004), and healthy (Hall, 2004) so on. But there are few researches about how these factors influence facial expression recognition. During facial expression recognition, the eye fixation location plays an important role. Fixation sites and fixation durations are closely time-locked to ongoing perceptual and cognitive process during picture perception (Henderson, 2005). We can understand the essence of facial expression recognition through analyzing the eye movement of facial expression recognition.
The aim of the research presented here was to explore the on-line process of facial expression recognition. This study consists of two experiments. The experiment 1 investigated the basic model of on-line process of recognizing the heterogeneity facial expression of undergraduate; the experiment 2 examined the effect of different facial part on recognition facial expression using shade technology.
Method
In two experiments, the recognition-decision paradigm was used. The participants were required to decide which kinds the facial expression pictures belong to- positive, neutrality and negative. Thirty-six undergraduate students participate in two experiments (18 in experiment 1 and 2 respectively). When participants performed, their eye movement was recorded by using 504 model eyetracker produced by America Science Laboratory. Excel managed the raw data. For testing significant differences between participants’ response to stimuli, ANOVA analyses were performed. The null hypothesis was rejected if a p-value less than 0.05 were obtained. All data were analyzed using SPSS 11.0 (SPSS inc, Chicago, USA).
Results
The data analyses showed that: (1) the eye trace of the heterogeneity expression recognition of subjects has common character which their eye moves the same path of "counter V"; (2) there is significant difference in behavior data and eye movement index when subjects recognized the heterogeneity expression recognition; (3) the pattern of eye movement have obviously changed when the diverse part of face were camouflaged. Defilade has influence on reaction time and accuracy data of facial expression recognition; (4) the information of different facial part has diverse effect on facial expression recognition. Eyes act as a most important role.
Conclusion
The results suggested that there is common character of eye movement model of the human individual recognizing the different facial expression. Psychological resources which needs during recognizing the heterogeneity facial expression are different. Moreover，facial expression recognition depended differently on different face parts. The eye part have most important role.

Key words: facial expression, recognition, on-line processing, eye movement