People use audio and visual signals to make sense of the world and to communicate within it.  Developing machines with the same (or enhanced) capabilities has been an exciting and challenging research theme in computer science for decades.  To accomplish this task, computers must process and generate audio and visual signals.

The Graphics, Vision and Speech Laboratory is concerned with the analysis, processing, recognition and generation of these signals in applications such as colour vision, machine vision, computer graphics, avatars and speech, music and language processing.  These technologies have many common theoretical foundations that include signal-processing, machine learning, statistical pattern recognition, time-series estimation, automata theory etc.

Although each area within the laboratory is an internationally-leading research group in its own right, there is collaboration and synergy between them in research in specific application areas.  For instance, our research on audio-visual speech synthesis and lip-reading draws on our experience in both automatic speech recognition and computer vision; research in automatic sign language generation integrates research in graphics, avatars and speech synthesis, and research in colour imaging is closely connected with computer vision. Our laboratories are equipped with fast graphics computers, haptic devices, spectrometers, 3D displays, a well equipped soundproof recording booth and an eight camera video-based motion capture system.

The laboratory is particularly successful in collaboration with industry: the laboratory dominates the University's patent portfolio, the University's spin-out portfolio, and the School's EPSRC grant and consultancy holdings.