2002 | OriginalPaper | Buchkapitel
Face Detection Based on Cost-Sensitive Support Vector Machines
verfasst von : Yong Ma, Xiaoqing Ding
Erschienen in: Pattern Recognition with Support Vector Machines
Verlag: Springer Berlin Heidelberg
Enthalten in: Professional Book Archive
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
This paper presents a method of detecting faces based on cost-sensitive Support Vector Machines (SVM). In our method, different costs are given to the misclassification of having a face missed and having a false alarm to train the SVM classifiers. The method achieves significant speed-ups over conventional SVM-based methods without reducing detection rate too much and the hierarchical architecture of the detector also reduces the complexity of training of the nonlinear SVM classifier. Experimental results have demonstrated the effectiveness of the method.