DocumentCode :
3412318
Title :
Facial expression recognition based on Local Sign Directional Pattern
Author :
Castillo, J.A.R. ; Rivera, Adin Ramirez ; Oksam Chae
Author_Institution :
Dept. of Comput. Eng., Kyung Hee Univ., Yongin, South Korea
fYear :
2012
fDate :
Sept. 30 2012-Oct. 3 2012
Firstpage :
2613
Lastpage :
2616
Abstract :
In this paper, we propose a novel local feature descriptor, Local Sign Directional Pattern (LSDP), for face expression recognition. LSDP encodes the directional information of the face´s textures-i.e., the texture´s structure-in a compact way, producing a more discriminating code than other state-of-the-art methods. The structure of each micro-pattern is encoded by using its prominent directions and sign-which allows it to distinguish among similar structural patterns that have different intensity transitions. We divide the face into several regions, from which we extract the distributions of the LSDP features. These features are concatenated into a feature vector, and used as a face descriptor, and the expression recognition is obtained with the aid of Support Vector Machine classifiers.
Keywords :
face recognition; feature extraction; image classification; image coding; image texture; support vector machines; LSDP feature extraction; directional information; discriminating code; face descriptor; face region; face texture; facial expression recognition; feature vector; intensity transition; local feature descriptor; local sign directional pattern; micropattern encoding; structural pattern; support vector machine classifier; texture structure; Accuracy; Databases; Face; Face recognition; Feature extraction; Support vector machines; Face descriptor; Local Sign Directional Pattern; Local patterns;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Image Processing (ICIP), 2012 19th IEEE International Conference on
Conference_Location :
Orlando, FL
ISSN :
1522-4880
Print_ISBN :
978-1-4673-2534-9
Electronic_ISBN :
1522-4880
Type :
conf
DOI :
10.1109/ICIP.2012.6467434
Filename :
6467434
Link To Document :
بازگشت