Monday 23 November 2026
6:00pm – 7:00pm

AI Translation between Sign and Spoken Language

13/14 Cornwall Terrace, Outer Circle (entrance facing Regent's Park), London NW1 4QP

Organised by the Daiwa Anglo-Japanese Foundation

AI now translates between many spoken languages, but sign languages have been largely left behind. In this talk, Professor Richard Bowden traces how AI has learnt to recognise, translate and produce sign language from video, from early research to today’s state of the art. He will show how this work is reaching the real world through Signapse, a University of Surrey spin-out whose AI signers are already used in settings such as public transport. The talk will discuss UMCS, a new five-year UK–Japan research project funded by EPSRC and the Japan Science and Technology Agency. Most sign language AI is trained on interpreted signing from television, which looks very different from how deaf people actually talk to each other. UMCS will instead focus on natural conversations in British Sign Language and Japanese Sign Language.

About the contributor

Professor Richard Bowden

Richard Bowden is Professor of Computer Vision and Machine Learning in the Centre for Vision, Speech and Signal Processing at the University of Surrey, where he leads the Cognitive Vision Group. His research uses computer vision and AI to locate, track and understand people, and he has been developing computational approaches to sign language recognition and translation for over 30 years. His group now focuses entirely on sign language technology. As well as Signapse, of which he is co-founder, his current projects include SignGPT, an EPSRC Programme Grant building the next generation of AI tools for sign languages; a Google.org grant to reduce the cost of sign language translation; and UMCS, a UK–Japan collaboration on real-time translation and learning across signed and spoken languages. He is a Fellow of the International Association for Pattern Recognition and a Distinguished Fellow of the British Machine Vision Association.