Mark Gales
Mark Gales is a researcher at Cambridge, where his work covers speech recognition, which turns audio into text, and speech synthesis, which turns text into audio. During his PhD he developed Parallel Model Combination, one of the first model-based compensation techniques for noise robustness, and he also proposed constrained MLLR, a transform that has become a standard transformation in deep learning frameworks. He returned to Cambridge as a University Lecturer in 1999, and he received IEEE Fellow status in 2011. He was elected an ISCA Fellow in 2019 for contributions to speech recognition, synthesis and statistical modelling algorithms. As of October 2026, he is a Research Professor in Information Engineering, and technology from his group is commercially deployed for tests such as Linguaskill.
- Now at
- University of Cambridge
- Works
- 567
- Citations
- 15,233
- h-index
- 53
Publication numbers from OpenAlex, as of 15 October 2026.
Timeline
-
1985–1988
First speech recognition system
He studies for a BA in Electrical and Information Sciences at Cambridge from 1985 to 1988, and in 1988 he builds his first automatic speech recognition system for his final-year project.
-
1995
A PhD on robust recognition
He completes his doctoral thesis on robust speech recognition in 1995. During that doctorate he develops Parallel Model Combination, one of the first model-based compensation techniques for noise robustness.
-
1995
A research fellowship in engineering
He starts working with Phil Woodland in Cambridge's Engineering Department as a Research Fellow in 1995. He proposes constrained MLLR (CMLLR), a transform that later becomes a standard transformation in deep learning frameworks.
-
1999
Return to Cambridge as lecturer
He returns to Cambridge's Engineering Department as a University Lecturer in 1999, when he is also appointed to the Emmanuel Fellowship.
-
2011
IEEE Fellow status
He receives IEEE Fellow status in 2011 for his contributions to acoustic modelling for speech recognition.
-
2019
An ISCA Fellowship
He is elected an ISCA Fellow in 2019 for contributions to speech recognition, synthesis and statistical modelling algorithms.
-
As of October 2026
Research professor at Cambridge
As of October 2026, he is a Research Professor in Information Engineering, and his research covers speech recognition and speech synthesis. Technology from his group is commercially deployed for tests such as Linguaskill.
Sources and credits
Photo: We have no published photo of Mark Gales yet. Send us a free-licensed photo.