The face retrieval problem, known as face detection, can be defined as follows: given an arbitrary black and white, still image, find the location and size of every human face it contains. There are many applications in which human face detection plays a very important role: it represents the first step in a fully automatic face recognition system, it can be used in image database indexing/searching by content, in surveillance systems and in human-computer interfaces. It also provides insight on how to approach other pattern recognition problems involving deformable textured objects. At the same time, it is one of the harder problems in pattern recognition.
We've designed an inductive learning detection method that produces a maximally specific hypothesis consistent with the training data. Three different sets of features were considered for defining the concept of a human face. The performance achieved is as follows: 85% detection rate, a false alarm rate of 0.04 % of the number of windows analyzed and 1 minute detection time on a 320 x 240 image on a Sun Ultrasparc 1.