PhD defence: Andreas Søeborg Kirkedal
In order to obtain the PhD degree, Andreas Søeborg Kirkedal has submitted his thesis entitled:
Danish Stød and Automatic Speech Recognition
The dissertation is a quantitative study of the Danish prosodic feature 'stød' and the application of stød in automatic speech recognition.
The thesis examines the reliability of stød annotation in the context of speech technology and use the annotation to rank acoustic features according to how well they predict stød in speech.
Furthermore, the thesis explores several methods to estimate statistical models that detect stød in speech.
Finally, several implementations of stød annotation and stød-predictive features in speech recognition are explored and evaluated with a speech recognition system developed during the project.
Supervisor:
Associate Professor Peter Juel Henrichsen
Department of International Business Communication
Copenhagen Business School
Assessment Committee:
Associate professor, Ph.D. Laura Winther Balling (Chair)
Department of International Business Communication
Copenhagen Business School
Torbjørn Nordgård
CEO
LingIT A/S
Professor Simon King
School of Philosophy, Psychology and Language Sciences
University of Edinburgh
Thesis:
The thesis is available here.
Reception:
The Doctoral School of Language, Law, Informatics, Operations Management, Accounting and Culture will host a reception, which will take place immediately after the defence in DH 2V.071.