Fergus McInnesCentre for Speech Technology Research (CSTR)
School of Informatics
University of Edinburgh
10 Crichton Street, Edinburgh, UK
Opinions expressed on these pages are personal and do not represent any official position of the University of Edinburgh, the School of Informatics, or the Centre for Speech Technology Research.
I am currently working at CSTR on the speaker diarization and speech recognition system developed on the AMI project (Augmented Multiparty Interaction). The system was originally developed for recognition of meeting audio, and I am exploring its application to other types of data such as lectures and podcasts.
From March 2010 to March 2011 I was a visiting researcher at CSTR; my work during this period focused on the application of a pattern discovery algorithm (developed from the segmental DTW algorithm of Park and Glass) to a corpus of infant-directed speech, to simulate the process that a child performs in the early stages of learning words.
From 2000 to 2009 my work was mostly on dialogue design, experimental design and statistics for applied human-computer interface projects in the Centre for Communication Interface Research (CCIR) - in particular the Spotlight project (2000-2) on spoken dialogue engineering for financial services and travel applications, and telephone dialogues research for Lloyds TSB (1998-2009) and for Standard Life Bank (2002-3).
I acted as second supervisor for the following recent PhD students within CCIR:
I also did some work on quality of service for voice over IP with
Nestor Becerra Yoma of the University of Chile during 2000 and 2001.
Education and past work