Visual aesthetics of a person is the set of visual preferences that makes an image appear more favorably. Recently, research emerged that utilized visual aesthetic cues of favorite images for person identification. All recent methods have relatively low identification rates (below 73%) due to the lack of discriminative visual features. In this thesis, we resolve this problem and introduce an efficient person identification method that achieves 84% identification rate. We also introduce a discriminative visual pattern and propose a novel method for gender recognition using a person's favorite images. The proposed gender recognition method demonstrates 91.20% accuracy in distinguishing males' and females' aesthetics on a Flickr database. Applications of
thesis findings to biometric research, forensics and recommender systems are discussed in the concluding section.