Research projects (CV)

CLARIN-NL project

NWO Small Programme (NWO Klein Programma)

Stevin demonstratie project 

Title: Development and Integration of Speech technology into Courseware for language learning
On using speech technology in CALL applications for learning pronunciation, morphology, and syntax.

Title: 'Implications of potentially erroneous feedback in CALL systems'.
Automatically generated feedback will, inevitably, contain errors. The major objective of this project is to advance our understanding of the effects on learning of the frequency and the seriousness of feedback errors in CALL systems.

On April 19th, 1999 the Dutch 'Platform voor het Nederlands in taal- en spraaktechnologie (TST)' was set up. The main goal of this Platform is to advance the position of the Dutch language within Language and Speech Technology.

The aim of this feasibility study is to determine how well ASR works for dysarthric speech. This study is done under the authority of, and in cooperation with the Sint Maartenskliniek Nijmegen.

MUMIS: Multimedia Indexing and Searching Environment
FP5: 5-th framework programme
MUMIS will develop basic technology for automatic indexing of multimedia programme material. In this project our main contribution is automatic transcript generation for the audio signals of radio and TV broadcasts.

In 1997 2 Ph.D. projects started of which I'm the supervisor:
1. Mirjam Wester: Modeling pronunciation variation for ASR: the top-down approach.
2. Judith M. Kessens: Modeling pronunciation variation for ASR: the bottom-up approach.

(Official Dutch name of the project: IT-project ITU96010 'Computergestuurde Spreekvaardigheidstoets')
The aim of this Senter project was to determine to which extent automatic assessment of foreign speakers' oral proficiency in Dutch is feasible with ASR technology.

This post-doctoral project of the KNAW (Royal Dutch Academy of Arts and Sciences) was entitled: "Using Articulatory Knowledge in Automatic Speech Recognition". The goal of this project was to improve the performance of continuous speech recognizers by using phonetic/phonological knowledge, especially knowledge about articulation. More information can be found here.

PP-TST: 'Prioriteits-Programma Taal- & Spraak-Technologie'
The Dutch National R&D programme 'Language and Speech Technology' of the Netherlands Organization for Scientific Research (NWO). The practical goal of this project was to build a working demonstrator of an interactive spoken language system that can give travel advice by telephone to individuals using public transport in the Netherlands.

MAIS : Multilingual Automatic Inquiry Systems
MLAP-94 : Multilingual Action Plan 1994
LE : Language Engineering
The goal of the MLAP-94 project MAIS was to transfer an existing German spoken dialogue system to French and Dutch. The original German system was developed by Philips Research Aachen (PFA : Philips Forschung Aachen), and it provided train time-table information via the telephone. The resulting Dutch system constituted the starting point of the NWO project PP-TST and the European ARISE project.

"Polderland": typically Dutch, but internationally-oriented. Why "Polderland"? This name refers to the technical ingenuity with which the Dutch have reclaimed land, the polders, which have made The Netherlands world-famous. Polderland's products and services distinguish themselves through the same degree of ingenuity and high quality. Our programs allow efficient and reliable automatic language and speech processing.

Since August 1992, I have worked at the Department of Language and Speech of the University of Nijmegen: I have carried out research on various topics, and I have been teaching phonetics, speech recognition and signal processing. Initially, the teaching was mainly done for two full-time MA-curricula of the Department of Language and Speech: Speech and Language Pathology & Language, Speech and Informatics.
The research first was a follow-up of the research done for my Ph.D. project mentioned above. On 28 November 1994, I successfully defended my Ph.D. thesis entitled: "Physiological control and behaviour of the voice source in the production of prosody".
Some specific research projects in which I was involved after that time are mentioned here. More information about my research topics can also be found here.

The ESPRIT (European Strategic Programme for Research and development in Information Technology) project POLYGLOT comprised research on speech recognition and speech synthesis for 7 languages of the European Community. The final goal of the project was a system with which one could communicate in 7 languages of the European Union: the system first recognized which language was spoken, and then switched to the recognition and synthesis modules for that specific language. Within this project I mainly did research on knowledge-based speech recognition for Dutch.

This three-year research project on the physiological control of intonation was carried out at the Department of Language and Speech of the University of Nijmegen. Part of this study was conducted at the Haskins Laboratories in New Haven.
Promotores: Prof. Dr. L.W.J. Boves. and Prof. Dr. C.C.A.M. Gielen
More information about my Ph.D. thesis can be found here.

From September 1984 till December 1985 I received practical training in order to obtain a (first degree) qualification to teach physics. Part of this training consisted of teaching physics in secondary school from August 1985 till December 1985.

For my M.Sc. thesis I did research on the dimensional and dynamical aspects of the auditory neurons in the midbrain (torus semicircularis) of the grassfrog (Rana temporaria L.), at the Department of Medical Physics and Biophysics of the University of Nijmegen.
Supervisors: prof. dr. J.J. Eggermont and Dr. W. Epping.