GEOFFREY WILLIAMS BSc, MA, PhD
Telephone (&fax): +44 (0)1245 261688
e-mail: geoff@phon.ucl.ac.uk
http://www.phon.ucl.ac.uk/home/geoff
CAREER HISTORY
Jan 2002 – June 05:
Senior Research Fellow, Dept of Phonetics & Linguistics, UCL
Senior researcher on SYNFACE (http://www.speech.kth.se/synface), a joint project with KTH, RNID and industrial partners to develop a synthetic talking face driven by speech recognition to provide speech-reading support for the hearing-impaired. Responsible for software development for perception experiments to evaluate benefits of system, training phonetic recognisers for Dutch and English, and investigation of speech recognition techniques for generating effective oral movements.
October 03 – Jan 05: WGNet project: software developer for second phase of an interactive computational model for Word Grammar (see http://www.phon.ucl.ac.uk/home/dick/wgnet/wgnet.htm).
[May 04 – Feb 05: seconded for 0.2 FTE as trade union negotiator on National Pay Framework (0.4 FTE from Feb)]
________________________________________________________________________________________________
March -Nov 2000:
Freelance software Consultant, Centre for Speech and Language Research, University of Colorado, USA
Devised and wrote Reading Tutor software primarily for deaf children, based on the CSLU Speech Toolkit, using Tcl/Tk under Windows 98/NT. Demonstrated program at Toolkit Short Course, INSTIL 2000 Conference, University of Abertay, Dundee, August 2000.
________________________________________________________________________________________________
Oct-Dec. 1999:
Lecturer (p-t) in Computational Linguistics, School of Oriental & African Studies, London,
Dept of Linguistics.
Taught and assessed an introductory under-/post-graduate course in natural language processing with Prolog, including lectures and lab classes. 5 hours per week contact time. Temporary post (one term).
________________________________________________________________________________________________
August 1992 - Feb. 1999:
Phonologist / Researcher, “Project HAL”. SOAS Linguistics Dept. and RMS Inc, Colorado, USA:
Commercially funded research in Automatic Speech Recognition: applying phonological theory (government phonology) to multi-lingual ASR. Supervised recording and transcription of a large multi-lingual speech database (German/Spanish/Japanese/Mandarin); performed statistical pattern recognition experiments (using SPSS); built a rule-based recogniser to demonstrate the viability of the approach (based on ESPS/waves). Also built standard polyphone speech recognisers using a standard package (HTK), and researched the viability of alternative recogniser architectures based on syllable structure and sub-segmental targets. Member of small team (maximum 4 at any one time).
Completed my PhD during this period, and contributed articles and conference papers. Filed a successful joint patent application in several countries. Relocated to RMS in Boulder, Colorado for final year of project.
________________________________________________________________________________________________
Feb 1991 - June 1992:
Research Assistant (RA 1A), SOAS, Department of Linguistics.
Main researcher on ESRC project Towards a Principle-Based Phonology: Theory and Application.
Job description as above, but more academically oriented research program. Basic research in 2-member team into acoustic properties of primitives of government phonology. Work resulted in 2 publications and some early results were incorporated into government phonology theory.
Taught exam revision classes in Phonology for the MA program in Linguistics, including marking assignments.
April 1989 - Jan 1991:
Computer Programmer/Advisor, SOAS, IT Department.
Support and advisory work in micro-computing and networking for School staff and students. Planned and installed the School's first Novell/Ethernet network. Involved in selection, recruitment and supervision of junior staff.
Senior member of department of 3.
_________________________________________________________________________________
Oct 1988 - April 1989:
Self-employed tutor / freelance translator.
Part-time Maths and English tutor (from elementary up to RSA stage 1) to adults on Employment Training Scheme. Combined this with freelance technical translation from German and French into English, mainly patents, with clients including RWS Group. ________________________________________________________________________________
Jan 1980 - Oct 1988:
Research Engineer, Marconi Research Centre, Gt Baddow, Essex.
Various departments: worked for 4 years on development of novel electro-mechanical sensors. Subsequently moved into Hybrid Circuits Group, where I was responsible for development of automatic test software for all in-house components.
________________________________________________________________________________________________
EDUCATION, ADDITIONAL SKILLS and EXPERIENCE
Higher Education:
1991-98 (part-time): PhD in Phonology and Speech Processing SOAS, University of London
Thesis title: “The phonological basis of speech recognition”.
Available from: http://www.phon.ucl.ac.uk/home/geoff/thesis.html
1989-91 (part-time): MA (Linguistics) (Distinction) SOAS, University of London
1976-79 (full-time): BSc (Electronics) Class 2.ii University of Southampton
Adult Education:
1987: Russian A-level (grade A) (Central London Polytechnic)
1983: German.: Institute of Linguists Intermediate Diploma, (Brentwood AEC, Essex)
School:
1976: Liverpool Institute: A-levels: Advanced Maths (B), Physics (B), Further Maths (D)
Academic Awards
· One year ESRC Postgraduate Research Award, 1996-97 for completion of PhD.
· Nominated for best Taught Masters prize for 1991, SOAS.
Programming Languages, Software Packages and Operating Systems
Unix, Dos/Windows:
· 10 years’ C programming under Unix, 1 year’s C++ under Windows. Competent in shell scripting (bash/tclsh); working knowledge of Perl, good knowledge of Prolog.
· 2 years’ experience of Tcl/Tk under various platforms. Working knowledge of LATex typesetting and HTML.
· Experienced user of HTK speech recognition package, NICO neural network package and various standard speech analysis and editing software, including ESPS/waves.
· Competent in SPSS for Windows, Microsoft Office packages, Visual C++.
· Experience in system administration for Linux, Solaris, Windows NT servers and Novell NetWare.
Foreign Languages
· German ‑ speak and read fluently. Professional experience of technical translation (patents, scientific papers)
· Russian, French ‑ reading knowledge, reasonable spoken fluency.
· Spanish, Mandarin Chinese - basic conversation.
Vocational and practical courses attended
· Various trade union courses related to Universities’ National Pay Framework and employment law.
· Basic Education for Employment Purposes: teacher training for tutors in Adult Basic Education (1988, 12 weeks)
Professional activities
Professional bodies:
International Speech Communication Association (2002-5)
International Phonetic Association, member (2003-4)
Acoustic Society of America, Student member (1996-97)
Insitute of Linguists, Associate Member (1983-1991)
Trade union:
Asoociation of University Teachers (AUT, now UCU):
Co-president of UCL AUT branch and part-time paid negotiator on pay and grading July 2004- June 05.
UCLAUT Exec committee member and case worker 2002-05;
UCLAUT delegate to AUT Council 2002-05.
Member of SOAS AUT executive committee 1994-98, negotiating on behalf of fixed-term contract staff, and Vice-President for 1 year.
Served on Chelmsford Trades Council as ASTMS/MSF delegate 1983-1986.
Voluntary:
I spent around 18 months as a volunteer adult literacy and numeracy tutor in 1980’s.
Member of Amnesty International, Campaign against the Arms Trade, and World Development Movement
PUBLICATIONS LIST for Geoff Williams, PhD
(Most of these are downloadable from http://www.phon.ucl.ac.uk/home/geoff/pubs)
Books & Articles:
PhD Thesis: The phonological basis of Speech Recognition. SOAS, University of London (1998).
Williams, G. (2003). On the computability of certain derivations in Government Phonology. In S. Ploch (ed) A Festschrift for Jonathan Kaye. Berlin: Mouton.
Siciliano, C., G. Williams, J. Beskow, & A. Faulkner (2003). Evaluation of a multilingual synthetic talking face as a communication aid for the hearing impaired. Speech Hearing & Language 14, UCL Dept of Phonetics & Linguistics.
Williams, G. (1997) A pattern-recognition model for the phonetic interpretation of elements. SOAS Working Papers in Linguistics, Vol 7, 1997.
Williams, G. (1994) On the role of phonological parsing in Speech Recognition. SOAS Working Papers in Linguistics, Vol 4, 1994.
Williams, G. and Brockhaus, W. (1992) Automatic Speech Recognition: a principle-based approach. SOAS Working Papers in Linguistics, Vol 2.
Non-linguistics:
Williams, G. & J.B. Flower (1988). Automated Testing of Hybrid Systems - strategy and implementation. GEC Review, Vol 4(1), 1988.
Conference Papers and talks:
Siciliano, C., Williams, Geoff, & A. Faulkner. Lip-readability of a Synthetic Talking Face in Normal Hearing and Hearing- Impaired Listeners. Proceedings of AVSP 2003: ISCA Tutorial and Research Workshop on Audio Visual Speech Processing 2003, St. Jorioz, France.
Siciliano, C., Williams, Geoff, Beskow, J. & A. Faulkner. Evaluation of a Multilingual Synthetic Talking Face as a Communication Aid for the Hearing Impaired. Proceedings of 15th ICPhS, Barcelona 2003; Session T5.5: Symposium on Applications of Auditory/Visual Speech Processing.
Williams, G., A.M. Terry and J.D. Kaye (1998). Phonological elements as a basis for language-independent ASR. Proceedings of ICSLP 98, Sydney Australia.
Williams, G., G. Martindale, A.M. Terry and J.D. Kaye (1996a). Multi-lingual speech recognition using phonological primes. Poster presented at ASA-AJA conference, Honolulu, December 1996.
Williams, G., G. Martindale, A.M. Terry and J.D. Kaye (1996b). Multi-lingual speech recognition using phonological primes. Proceedings of SPECOM 96, St Petersburg, Russia, October 1996. (sole presenter)
Williams, G. (1995). Speech recognition and the formal power of phonology. Presented at Computational Phonology Workshop, ESSLLI 95 Summer school, Barcelona, Spain. August 1995.
Williams, G. and J.D. Kaye (1994). For a principle-based approach to Speech Recognition: arguments from phonological structure. Proceedings of the Workshop on Computational Linguistics for Speech and Handwriting Recognition, Leeds University, England, 1994.
Williams, G. (1992). Automatic Speech Recognition: a Government Phonology approach. In Dressler et al (eds) Phonologica 1992: Proceedings of the 7th International Phonology Meeting. Krems, Austria. 1992. (refereed)
Automatic Speech Recognition: a Government-Based Approach. Presented at London Phonology Seminar, November 1991.
Patents:
Joint holder with Jonathan Kaye of US, European & Japanese patents for Phonology-Based ASR. (US Patent # 5,623,609)