Intelligent Face-Based Mobile Images Categorization

Given the increasing number of mobile platforms, a key technical challenge is how to provide an optimal photo browsing experience given the limited screen size available on mobile devices. This paper proposes a novel technique for intelligent mobile image categorization to reduce computation complexity on the Android mobile platform. In this technique, captured images are analyzed directly in JPEG format and then classified in real-time based on the gender of the human subjects appearing in the image. To increase system robustness in various light conditions, DC coefficients are discarded. In addition, to reduce complexity while effectively differentiating subject gender, a set of AC coefficients are automatically selected based on a three-step dimensionality reduction, in which evaluation of the coefficients’ significance is conducted by an LDA-based approach. Experimental results obtained using extensive datasets captured under uncontrolled conditions show the proposed system effectively manages photo on resourcelimited mobile platforms.