Mark Gales

Mark Gales is a Professor of Information Engineering in the Machine Intelligence Laboratory (formerly the Speech Vision and Robotics (SVR) group) and a Fellow of Emmanuel College. He is a member of the Speech Research Group together with faculty staff members Steve Young, Phil Woodland, Milica Gasic, and Bill Byrne.

A brief biography is available online.

Research interests

Low Resource Speech Processing - BABEL and MATERIAL projects
Automatic Assesment of Spoken English - The ALTA Institute
Uncertainty, Generalisation and Interpretability for Deep Learning - Google Research Award
High Quality Expressive, Controllable and Adaptable Speech Synthesis - HQSTS project
Large Vocabulary Speech Recognition - MGB Challenge (English)

A technical review of "traditional" HMM-based systems see The Application of Hidden Markov Models in Speech Recognition. For more general, and deep-learning based, approaches see this draft chapter Statistical Sequence Modelling.

If you are interested in studying for a PhD in Spoken Language Processing contact Mark Gales.

CUED RNNLM Toolkit (v1.1 release December 2017)
Percival TTS CNN-WGAN acoustic model for text-to-speech synthesis (HQSTS project)
HTK Version 3 - HTK_V3.4.1 and examples are available.
HTK_V3.5 beta updted 28th June 2016, see UK Speech Presentation for details

top

For the Engineering Department:

MPhil in Machine Learning Speech and Language Technology - local information
- Speech and Language Processing Applications (MLSALT5)
- Statistical Speech Synthesis (MLSALT10)
Fourth year engineering undergraduates Module 4F10: Deep Learning and Structured Data.
Fourth year engineering undergraduates Final Year Projects.
Third year engineering undergraduates Module 3M1: Mathematical Methods.

For Emmanuel College: (Information local access only)

top

Mark Gales
Baker Building, Room 305
Engineering Department	Email mjfg@eng.cam.ac.uk
Trumpington Street, Cambridge	Tel: +44 1223 332733
CB2 1PZ, UK	Fax: +44 1223 332662