TY - GEN
T1 - Face detection based on support vector machines
AU - Xi, Dihua
AU - Lee, Seong Whan
PY - 2002
Y1 - 2002
N2 - Face detection is a key problem in building an automatic face system such as face recognition and authentication. A number of approaches have been proposed for face detection. Recently, a novel statistical machine learning method, support vector machine, has been employed. Generally, the current SVM-based methods can be divided into two categories: component-based and whole face-based. It is difficult for the component-based method to extract the small face due to no enough information for each component exists. On the other hand, the whole face-based method is too much computationally expensive to build an effective system. In this paper we present a fast system named wavelet-SVM method to extract a wide range scales of faces from grey-scale images or color images with a preprocessing using a TSL color model. The system is not only accurate and effective, but also largely speeds the system up by applying a TSL B-G color model and multiresolution wavelet decomposition.
AB - Face detection is a key problem in building an automatic face system such as face recognition and authentication. A number of approaches have been proposed for face detection. Recently, a novel statistical machine learning method, support vector machine, has been employed. Generally, the current SVM-based methods can be divided into two categories: component-based and whole face-based. It is difficult for the component-based method to extract the small face due to no enough information for each component exists. On the other hand, the whole face-based method is too much computationally expensive to build an effective system. In this paper we present a fast system named wavelet-SVM method to extract a wide range scales of faces from grey-scale images or color images with a preprocessing using a TSL color model. The system is not only accurate and effective, but also largely speeds the system up by applying a TSL B-G color model and multiresolution wavelet decomposition.
UR - http://www.scopus.com/inward/record.url?scp=23044534799&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=23044534799&partnerID=8YFLogxK
U2 - 10.1007/3-540-45665-1_29
DO - 10.1007/3-540-45665-1_29
M3 - Conference contribution
AN - SCOPUS:23044534799
SN - 354044016X
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 370
EP - 387
BT - Pattern Recognition with Support Vector Machines - First International Workshop, SVM 2002 Niagara Falls, Canada, August 10, 2002 Proceedings
A2 - Lee, Seong-Whan
A2 - Verri, Alessandro
PB - Springer Verlag
T2 - 1st International Workshop on Pattern Recognition with Support Vector Machines, SVM 2002
Y2 - 10 August 2002 through 10 August 2002
ER -