printer friendly version of this page

Spoken Language Processing by Mind and Machine

Tutorial at INTERSPEECH 2007, Antwerp, Belgium

This tutorial will compare and contrast spoken language processing as performed by machines with the corresponding processes performed by human beings.  Theories of human speech perception, production, cognition and discourse will be discussed alongside algorithms for automatic speech recognition, synthesis, understanding and dialogue.  Attention will be given to the latest research attempts to unify these areas within a common framework.

Great strides have been made in our understanding of how human beings use and process speech as well as in our ability to create and implement practical applications that incorporate voice-based interaction. However, our current level of knowledge is quite modest in comparison to the advanced communicative skills of the average human being, and the full potential of truly ubiquitous SL technology may not be able to be realised using today's models and algorithms.  It is therefore timely for the different research communities to develop a modest understanding of each other's progress.

Topics to be covered include:

• speech recognition & interpretation
• speech generation & synthesis
• discourse & dialogue
• individuality & emotion
• evolution of spoken language
• spoken language acquisition
• plasticity
• towards a unified view

Overall aims of the tutorial:

• to bridge the gap between human and machine spoken language processing
• to provide a general overview of the current research issues in spoken language processing by mind and machine

Learning outcomes:

• a basic understanding of the similarities and differences between human and machine spoken language processing
• an awareness of the key research issues

Presenters

Roger K. Moore,
Univ. of Sheffield

Anne Cutler
Max-Planck-Institute for Psycholinguistics, Nijmegen

Short Bios

Roger Moore studied Computer and Communications Engineering at the University of Essex and was awarded the B.A. (Hons.) degree in 1973.  He subsequently received the M.Sc. and Ph.D. degrees from the same university in 1975 and 1977 respectively, both theses being on the topic of automatic speech recognition.  After a period of post-doctoral research in the Phonetics Department at University College London, he was head-hunted in 1980 to establish a speech recognition research team at the Royal Signals and Radar Establishment (RSRE) in Malvern.  In 1985 he became head of the newly created ‘Speech Research Unit’ (SRU) and subsequently rose to the position of Senior Fellow (Deputy Chief Scientific Officer – Individual Merit) in the ‘Defence and Evaluation Research Agency’ (DERA).  Following the privatisation of the SRU in 1999, he continued to provide the technical lead as Chief Scientific Officer at 20/20 Speech Ltd. - a joint venture company between DERA (now QinetiQ) and NXT plc.  Since 2004 he has been Professor of Spoken Language in the ‘Speech and Hearing’ Research Group (SPandH) at Sheffield University.  Prof. Moore has authored and co-authored over 100 scientific publications in the general area of speech technology applications, algorithms and assessment.  He is a Fellow of the Institute of Acoustics and a Visiting Professor in the Department of Phonetics and Linguistics at University College London.  He is also a member of the editorial/advisory boards for the scientific journals ‘Computer Speech and Language’ and ‘Speech Communication’. He is past Chairman of the ‘European Expert Advisory Group on Language Engineering Standards’ (EAGLES) working party on spoken language resources, and Editor of the 'Handbook of Standards and Resources for Spoken Language Systems'.  Prof. Moore served as President of the ‘International Speech Communication Association’ (ISCA) from 1997 to 2001 and President of the Permanent Council of the ‘International Conferences on Spoken Language Processing’ (ICSLP) from 1996 to 2001. In 1994 he was awarded the prestigious UK Institute of Acoustics Tyndall medal for "distinguished work in the field of speech research and technology" and in 1999 he was presented with the NATO RTO Scientific Achievement Award for "repeated contribution in scientific and technological cooperation".

Anne Cutler is a director of the Max Planck Institute for Psycholinguistics in Nijmegen, the Netherlands, Professor of Comparative Psycholinguistics at the Radboud University Nijmegen, and Professor in MARCS Auditory Laboratories at the University of Western Sydney. She studied in Australia, Germany and the US, and before moving to Nijmegen she worked in the UK, chiefly with the Medical Research Council in Cambridge. Her research centres on how language-specific phonological structure affects the recognition of spoken language. Recent research has included studies of the recognition of Dutch, English, German, French, Japanese, Korean, Finnish, Cantonese, Sesotho, Spanish, Telugu and Arabic.