Multiple Instance Learning Based Adaboost Algorithm And Its Application In Face Detection

Human face is a normal and complicated vision mode, and the vision information on faces greatly influences the communication between persons. The disposal and analysis of faces has far-range application foreground in video monitoring, gate controlling, teleconferencing and human-computer interaction field, et al. Face detection is viewed as a special case of object detection, gets more and more researchers'attention. Face detection is the first and key step of face information disposal problem, the precision of face detection algorithm directly affects the following images'disposal and analysis steps.Due to the complexity of human face images, it's hard to describe the face feature obviously. So, Image-based face detection attracts many researchers'attention now. What the basic idea of image-based face detection is that it views face detection as a general pattern recognition problem. It classifies examples into two kinds: faces and non-faces by training all the examples.Adaboost algorithm is the most efficient image-based face detection approach which has high detection accuracy. However, what the problem...