Computing

Biographical Details:

Dr. John McKenna has been a Lecturer in Speech Technology and Computational Linguistics in the School of Computing at Dublin City University since September 2000. He currently lectures on Speech Processing (4th yr, undergrad), and Biometric (taught MSc. module), and has lectured in Statistical Techniques in Natural Language Processing, and Introductory Phonetics and Phonology . He graduated with the degree of B.Eng. in Electronic Engineering in 1988 from NIHE, Limerick (now UL). He also obtained the degrees of M.Sc. in Speech and Language Processing and Ph.D. titled "Kalman Filtering Towards Automatic Speaker Characterisation" from the University of Edinburgh in 1996 and 2004 respectively. While a Ph.D. student in Edinburgh he was involved in teaching a variety of subjects in the departments of Linguistics, Computer Science, and Artificial Intellingence. From 1988-1995, he worked as a design engineer (Electronics) in the London area and as a communications consultant in Seoul, Korea. He spent a year (1992/3) in voluntary teaching at QingHai University in the People's Republic of China. His research is primarily focused on the problem of Speaker Characterisation and he is particularly interested in probabilistic modelling and machine learning approaches to the processing of spoken language. He is currently supervising two Ph.D. students ("Machine Learning of Speaker Characteristic Speech Dynamics and Interaction" and �Probabilistic Filtering and Coding of Speech�). He maintains close ties with and makes regular visits to the Centre for Speech Technology Research in Edinburgh, where he completed his Ph.D. studies. He is a member of the National Centre for Language Technology (NCLT) and RINCE, both based at DCU, and is a member of the International Speech Communication Association (ISCA) and the Irish Speech Group (ISG). He is currently CA1 year head and also sits on a number of committees internal to DCU. He has been internal examiner to 3 PhD theses within the School of Computing, and external examiner to one MSc thesis.

Research Interests:

Speech Technology: Speaker Characterisation Stochastic modelling and data-driven learning of speech and language patterns. Speech Technology in CALL (Computer Assisted Language Learning)

Selected Books

• John G. McKenna. Kalman Filtering Towards Speaker Characterisation, Unpublished, 2001.

• Errity, A. and McKenna, J., A comparison of linear and nonlinear dimensionality reduction methods applied to synthetic speech, Interspeech 2009 - Eurospeech, 06-SEP-09 - 10-SEP-09, Brighton, UK
• Errity, A., McKenna, J., and Kirkpatrick, B., Dimensionality reduction methods applied to both magnitude and phase derived features, Interspeech 2007 - Eurospeech, 26-AUG-07 - 29-AUG-07, Antwerp, Belgium
• Errity, A. and McKenna, J, A comparative study of linear and nonlinear dimensionality reduction for speaker identification, 15th Int. Conf. on Digital Signal Processing (DSP), 01-JUL-07 - 04-JUL-07, Cardiff, Wales
• Errity, A., McKenna, J., and Kirkpatrick, B., Manifold learning-based feature transformation for phone classification, ISCA Tutorial and Research Workshop on Nonlinear Speech Processing (NOLISP), 22-MAY-07 - 25-MAY-07, Paris, France
• Errity, A. and McKenna, J, An investigation of manifold learning for speech analysis, Intl. Conf. on Spoken Language Processing (Interspeech 2006 - ICSLP), 17-SEP-06 - 21-SEP-06, Pittsburgh PA, USA.
• Tooher, M. and McKenna, J. G., Application of Glottal Source Information to Speech Recognition, SPECOM 2005, 10th International Conference on Speech and Computer, 17-OCT-05 - 19-OCT-05, University of Patras, Patras, Greece
• Errity, A., McKenna, J. and Isard, S., Unscented Kalman Filtering of Line Spectral Frequencies, 8th International Conference on Spoken Language Processing, 04-SEP-04 - 08-SEP-04, Jeju, Korea
• Tooher, M. and McKenna, J., Prediction of the Glottal LF Parameters using Regression Trees, 8th International Conference on Spoken Language Processing, 04-SEP-04 - 08-SEP-04, Jeju, Korea
• Tooher, M. and McKenna, J. G., Variation of glottal LF parameters across F0, vowels and phonetic environment, ISCA Tutorial and Research Workshop on Voice Quaility Functions, Analysis and Synthesis (VOQUAL '03), 27-AUG-03 - 29-AUG-03, Geneva
• John G. McKenna, Automatic Glottal Closed-Phase Location and Analysis by Kalman Filtering, 4th ISCA Tutorial and Research Workshop on Speech Synthesis, 29-AUG-01 - 01-SEP-01, Pitlochry, Scotland
• John McKenna and Stephen Isard, Tailoring Kalman Filtering Towards Speaker Characterisation, Eurospeech 99, 06-SEP-99 - 10-SEP-99, Budapest, Hungary