The unique research area of audio-visual speech recognition has attracted much interest in recent years as visual information about lip dynamics has been shown to improve the performance of automatic speech recognition systems, especially in noisy environments.