A NEW APPEARANCE-BASED FACIAL EXPRESSION RECOGNITION SYSTEM WITH EXPRESSION TRANSITION MATRICES

作者：Chang Yang Kai^*; Lien Cheng Chang; Lin Lin Yi

来源：International Journal of Innovative Computing Information and Control, 2009, 5(11B): 4089-4104.

• Facial expression recognition
• Expression transition
• Boosted tree classifiers

Generally, the facial expression recognition systems can be roughly categorized into feature-based, image-based and model-based methods. However, several problems exist in the above methods. First, most of feature-based methods can not extract the facial features (shape, color and position) robustly because of hair and glasses occlusion, wrinkle or illumination variation. Second, the computation of extracting facial features is complex and costly. Third, most of current researches in the three kinds of methods can't recognize the facial expressions at low-resolution images. To overcome these problems, a novel appearance-based facial expression recognition method called "expression transition' is proposed to identify six kinds of facial expressions (anger, fear, happiness, neutral, sadness and surprise) at low-resolution images efficiently. The boosted tree classifiers, Hough transform and template matching are used to locate and crop the effective facial region that may characterize the facial expressions. Then, the expression transformed images via a set of expression transition matrices are matched with the real facial images to identify the facial expressions. The proposed system can recognize the facial expressions with the speed of 0.24 seconds per frame and accuracy above 86%.

• 出版日期2009-11