Training convolutional filters for robust face detection

We present a novel face detection approach based on a convolutional neural architecture, designed to de tec t and precisely localize highly variable face pa t te rns , i n complex real world images. Our system automatically synthesizes s imple problem-specific feature ext rac tors f rom a training set of face and non face patterns, without making any assumptions or using any hand-made design concerning the features to extract or the areas of the face pattern to analyze. Exper iments on different difficult test sets have shown that our approach provide superior overall detection results, while being computat ionnal ly more efficient than most of state-of-the-art approaches that require dense scanning and local preprocessing.