A new model for face detection in cluttered backgrounds using saliency map and C2 texture features

Abstract

This paper introduces a new model for face detection, which is based on the saliency map of a visual attention model and C₂ features of the visual cortex HMAX model. The visual attention model is tuned by top-down skin color features for finding saliency maps of face candidate regions, and the top-down C₂ texture features of the visual cortex HMAX model are used for face detection task. After the candidate regions of faces are found, the C₂ texture features from them are extracted to categorize face or non-face regions using support vector machine classifier. A Caltech face database with background is utilized for testing the proposed model. Experimental results demonstrate that the proposed model provides plausible performance and it is reliable against variations in face illuminations, expressions and cluttered backgrounds.