Singular Value Feature and Hidden Markov Models-Based Face Detection
基于奇异值特征和隐马尔可夫模型的人脸检测

This paper presents a novel face detection algorithm, which consists of two parts of research work. The first one is a frontal view upright face detection algorithm which is based on the well known singular value feature (SVF) and hidden Markov models (HMM). A trained HMM is employed to classify the SVF of the sub-image at every location in the image as a face or nonface. The algorithm couples the virtues of both the SVF and HMM and produces excellent detection results. It is tested on a collect photo album and has detected the 85.1 percent of its 484 people, while 97 false alarms are also reported. The second part of our algorithm is the extension of the first one to rotation invariant face detection. Several HMMs are utilized to recognize the SVF of the sub image at the same time to obtain the angle of the "face" image. Then the HMM for detecting the upright faces is employed to verify the faceness of the rotated test pattern. The rotation invariant algorithm is tested on another image set where there are 173 persons. The detection rate is 72 2%,which seems not too high,but the false alarm rate is also low, only 34 out of the 6 100 000 windows,scanned.