speaker recognition, eigenfaces, feature extraction, user identification
In this paper, considerations and methodology for user identification in gaming applications is discussed. Typical user identification processes operate through input devices that yield qualitative audio and video data in a controlled environment. In gaming applications, low cost hardware and uncontrolled environmental conditions pose a serious obstacle towards efficient user identification. Through a combination of several methods, including speaker recognition, facial feature extraction and the eigenface recognition approach for face recognition, a robust algorithm can be developed. Weighing the relative importance of each method leads to more robust recognition, despite the limitations associated with the application.