Computer Science · Computer Vision and Pattern Recognition
Face Recognition System Using Gabor Features and HTK Toolkit
Zineb Elgarrai, Othmane El Meslouhi, Hakim Allali, Mustapha Kardouchi, Sid‐Ahmed Selouani
See on the mapAbstract
This paper presents a new face recognition system. The proposed system is built on Hidden Markov Models (HMMs). Facial image features are extracted using Gabor filters. The dimensionality of those features is reduced using the Linear Discriminant Analysis (LDA) method to keep only the most relevant information. Then, the system injects the resulting feature vectors to the Hidden Markov Model Toolkit (HTK). Note that HTK is a portable toolkit for speech recognition system. Experimental results on YALE and ORL databases show the efficiency of the proposed approach.