An efficient face image retrieval system based on attribute sparse codewords

Face recognition system is used to identify people from a digital image, which can be done by comparing selected facial attributes from the image and a face image database. Face image recognition systems used present variations in image quality, pose, illumination and resolution which help to improve the accuracy. In this paper we present attribute enhanced sparse codeword based face recognition and retrieval system which provide a better recognition performance when compare to the existing methods. Here the face is detected by the Active Shape Model Algorithm. It is good founded to uncontrolled pose images. Active Shape Models (ASMs) are statistical model of the shape of objects which iteratively deform to fit to an example of the object in a new image. We investigate the effectiveness of different attributes and important factors essential for face retrieval. By using LBP feature extraction model we are able to find better feature representations and achieve better retrieval results along with attribute sparse codewords. Experimenting on public datasets, the results show that the proposed method can achieve up to 43.4% relative improvement compared to the existing methods.