[Univ of Cambridge] [Dept of Engineering]

Catherine Breslin


I am currently a research associate in the Dialogue Systems group run by Prof Steve Young in the Machine Intelligence Laboratory at Cambridge University Engineering Department.

Prior to joining the group, I studied Engineering and Computer Science at Oxford University, graduating in 2003, before completing an MPhil in Computer Speech, Text and Internet Technology at Cambridge University in 2004. I undertook my PhD under the supervision of Dr Mark Gales in 2004-08, on generating and combining complementary systems for automatic speech recognition, and subsequently worked for Toshiba Research until 2012.

I instituted the Friday speech group seminars within the speech group.

View Catherine Breslin's profile on LinkedIn

Current project and research interests

  • Online automatic speech recognition (ASR) and voice activity detection (VAD) for dialogue systems
  • Combining VAD and ASR for improved performance
  • End-of-turn detection for improved turn-taking in dialogue
  • Incremental recognition and adaptation for more flexible dialogue management

Publications

Conference Papers

  • C. Breslin, M. Gašić, M. Henderson, D. Kim, M. Szummer, B. Thomson, P. Tsiakoulis and S. Young
    "Continuous ASR for Flexible Incremental Dialogue"
    ICASSP 2013, to appear
  • M. Gašić, C. Breslin, M. Henderson, D. Kim, M. Szummer, B. Thomson, P. Tsiakoulis and S. Young
    "On-line Policy Optimisation of Bayesian Spoken Dialogue Systems via Human Interaction"
    ICASSP 2013, to appear
  • C. Breslin, K.K. Chin, M.J.F. Gales and K. Knill
    "Integrated Online Speaker Clustering and Adaptation" (pdf)
    Interspeech 2011
  • K.K. Chin, H. Xu, M.J.F. Gales, C. Breslin and K. Knill
    "Prior Joint Speaker and Noise Compensation for Robust Speech Recognition" (pdf)
    ICASSP 2011
  • C. Breslin, H. Xu, K.K. Chin, M.J.F. Gales and K. Knill
    "Prior Information for Rapid Speaker Adaptation" (pdf)
    Interspeech 2010
  • C. Breslin, M. Stuttle and K. Knill
    "Compression Techniques Applied to Multiple Speech Recognition Systems" (pdf)
    Interspeech 2009
  • C. Breslin and M.J.F. Gales
    "Building Multiple Complementary Systems using Directed Decision Trees" (ps|pdf)
    Interspeech, August 2007
  • C. Breslin and M.J.F. Gales
    "Complementary System Generation Using Directed Decision Trees" (ps|pdf)
    ICASSP, April 2007
  • C. Breslin and M.J.F. Gales
    "Generating Complementary Systems for Speech Recognition" (ps|pdf)
    Interspeech, Sept 2006

Journal Papers

  • C. Breslin and M.J.F. Gales
    "Directed Decision Trees for Generating Complementary Systems" Speech Comunication May 2009

Patents

  • "Speech processing system and method" filed March 2011
  • "An adaptive speech recognition system and method using a cascade of transforms" filed May 2010
  • "An adaptive speech processing system" filed May 2010
  • "A speech recognition based method and system for retrieving data" filed November 2008 (secondary inventor)
  • "A speech recognition system using a plurality of acoustic models which share probability distributions" filed November 2008

PhD project (2004-2008)

  • Complementary model generation for automatic speech recognition
  • Minimum Bayes Risk Training
  • Model combination for Automatic Speech Recognition

PhD Thesis

  • C. Breslin (2008)
    "Generation and Combination of Complementary Systems for Automatic Speech Recognition" (pdf)
    Completed June 2008

Contact Information

Catherine Breslin
Engineering Department
Trumpington Street, Cambridge
CB2 1PZ, UK e-mail cb404(at)eng.cam.ac.uk

[ Cambridge University | CUED | MIL ]