Efficient Image Retrieval Using Sparse Codewords

Social network services such as facebook, twitter provide photo sharing facilities. Among all those photos majority photos contain human faces. People are interested to discover contents in images. Human beings are involved in uploading images, searching images in it. Hence with the exponentially growing photos, large-scale content-based face image retrieval is a facilitating technology for many emerging applications. Image retrieval is done using sparse codewords method. It is possible to build the system which automatically detects human attributes. Automatically detected human attributes are used to increase the performance of content based face image retrieval. Low level features of image are combined with high level features of image to acquire more efficient results of image retrieval. Semantic codewords are used for face image retrieval. Sparse codewords are constructed using semantic cues of the face image to have enhanced content based face image retrieval.