Kris Demuynck

Kris Demuynck completed his PhD on “Extracting, Modelling and Combining Information in Speech Recognition” in 2001 at the KU Leuven. His principal research interests are large vocabulary continuous speech recognition, machine learning and search algorithms. His PhD work and the subsequent research as post-doctoral researcher at the KU Leuven resulted, amongst others, in a speech recognition toolkit which was made available publicly under the name SPRAAK in 2008.

In 2011, he moved to Ghent where he works as part-time research manager at imec-IDLab-UGent and part-time Professor at Ghent University, Faculty of Engineering and Architecture. The research team of Kris Demuynck works on speech and language technology such as speech recognition (verbal information), speech diarisation (who speaks when + language & dialect recognition), extraction of paralinguistic (non-verbal) information such as emotion and mental state of the speakers, automatic assessment of speech, and generic audio processing. Some of the topic on which Kris Demuynck has worked and that are relevant to this human centric AI: algorithms to extract information from speech and audio (prosody, speaker identification, emotion and mental state …); deep learning (MLPs, CRFs and reservoir computing networks, …); speech assessment (reading tutor for children, tools for speech therapists); and most aspects of speech recognition including various human-inspired approaches.

Keywords: Speech Analysis, Audio Analysis, Text Analysis, Deep learning, Machine learning

Key publications
  • T. N. Sainath, B. Ramabhadran, D. Nahamoo, D. Kanevsky, D. Van Compernolle, K. Demuynck, J. F. Gemmeke, J. R. Bellegarda, and S. Sundaram, “Exemplar-Based Processing for Speech Recognition: An Overview”. In IEEE Signal Processing Magazine, vol. 29, no. 6, pages 98–113, Nov. 2012.
  • K. Demuynck, J. Duchateau, D. Van Compernolle and P. Wambacq, “An Efficient Search Space Representation for Large Vocabulary Continuous Speech Recognition”. Speech Communication, vol. 30, no. 1, pages 37–53, Jan. 2000.
  • F. Triefenbach, A. Jalalvand, K. Demuynck and J-P. Martens, “Acoustic modeling with hierarchical reservoirs”, In IEEE Transaction on Audio, Speech and Language Processing, vol. 21, no. 11, pages 2439–2450, Nov. 2013.
  • B. Desplanques, K. Demuynck, and J-P. Martens, “Adaptive speaker diarization of broadcast news based on factor analysis”, In Computer Speech and Language, vol. 46, pages 72–93, 2017.
  • Joris Pelemans, Kris Demuynck, and Patrick Wambacq. A layered approach for Dutch large vocabulary continuous speech recognition. In Proc. International Conference on Acoustics, Speech and Signal Processing, pages 4421–4424, Kyoto, Japan, March 2012.
Publication links