Home page
Call for papers
Expression of interest
Author's kit
Paper Submission
Scientific Committee
Technical program
Travel Information

Local pages
TLP home page
LIMSI home page
CNRS home page


Paris time: 16h09

Technical Program

All presentations will be oral. A time slot of 20 minutes (including questions) is allocated for regular papers.
A room will be available for demonstrations during the breaks. If you are interested in presenting a demo, please contact the workshop organizers at


Accepted papers

Full program (papers can be downloaded by clicking on the title):
Sept 18, 2000
Sept 19, 2000
Sept 20, 2000
9:30 - 13:00
Sessions 1-2
starting 11:00
Sessions 6-7
Sessions 10-11
13:00 - 14:30
14:30 - 18:30
Sessions 3-5
Sessions 8-9
Sessions 12-13
end 19:30
Welcome reception
Conference dinner

Monday Morning
10:00 Registration Open
11:00 Welcome
Keynote, 11:30-12:20
11:30 Large Scale Discriminative Training for Speech Recognition 
P.C. Woodland, D. Povey, Cambridge Univ. Engineering Dept.
Acoustic Model Training, 12:20-13:00
12:20 Combination and Joint Training of Acoustic Classifiers for Speech Recognition
Katrin Kirchhoff, Jeff Bilmes, Univ. of Washington, Dept. of Electrical Engineering
12:40 Tied-Posteriors: A New Hybrid Speech Recognition Technology with Generic Capabilities and High Portability 
Jan Stadermann, Jörg Rottland, Gerhard Rigoll, Gerhard-Mercator-Univ. Duisburg
13:00-14:30 Lunch

Monday Afternoon
3 Decoding, 14:30-15:50
14:30 The block-synchronous search algorithm
Peter Stubley, Locus Dialogue
14:50 A Decoder for Finite-State Structured Search Spaces
Diamantino Caseiro, Isabel Trancoso, INESC IST
15:10 Within-Word vs. Across-Word Decoding for Online Speech Recognition
Stephan Kanthak, Achim Sixtus, Sirko Molau, Hermann Ney, Lehrstuhl fuer Informatik VI, RWTH Aachen Univ. of Technology
15:30 Improved ROVER Using Language Model Information 
Holger Schwenk, Jean-Luc Gauvain, LIMSI-CNRS
15:50-16:30 Coffee break
4 Noise Robustness I, 16:30-17:30
16:30 Reliable ASR Based on Unreliable Features
I. Potamitis, N. Fakotakis, G. Kokkinakis, Wire Communication Lab., Univ. of Patras
16:50 Constrained Spectrum Normalization for Robust Speech Recognition in Noise
Filipp Korkmazskiy, Frank K. Soong, Olivier Siohan, Bell Labs Lucent Technologies
17:10 Noise Level Normalization and Reference Adaption for Robust Speech Recognition (slides)
Florian Hilger, Hermann Ney, Lehrstuhl fuer Informatik VI - RWTH Aachen Univ. of Technology
5 Language and Pronunciation Modeling, 17:30-18:30
17:30 Delimited smoothing technique over pruned and not pruned syntactic language models: perplexity and WER
A. Varona, I. Torres, Dpto. Electricidad y Electrónica, Univ. del Pais Vasco/EHU
17:50 Theory and Practice of Acoustic Confusability
Harry Printz, Peder Olsen, IBM Watson Research Center
18:10 Data-driven pronunciation modelling for non-native speakers using association strength between phones
Ingunn Amdal, Filipp Korkmazskiy, Arun C. Surendran, Multimedia Communications Research Laboratory, Bell Labs, Lucent Technologies
18h30 Welcome reception

Tuesday Morning
6 Keynotes, 9:30-11:40
9:30 A Brief Overview of Decoding Techniques for Large Vocabulary Continuous Speech Recognition 
Xavier Aubert, Philips Research Laboratories
10:20-10:50 Coffee break
10:50 Weighted Finite-State Transducers in Speech Recognition
Mehryar Mohri, Fernando Pereira, Michael Riley, AT&T Labs Research
7 Acoustic Model Adaptation, 11:40-13:00
11:40 New Adaptation Techniques For Large Vocabulary Continous Speech Recognition
Yuqing Gao, Bhuvana Ramabhadran, Michael Picheny, IBM T.J.Watson Research Center
12:00 Bayesian Adaptation Revisited
Patrick Kenny, Gilles Boulianne, Pierre Dumouchel, Centre de recherche informatique de Montreal
12:20 Structural Maximum a Posteriori Linear Regression for Fast HMM Adaptation
Olivier Siohan, Tor Andre Myrvoll, Chin-Hui Lee, Bell Labs, Lucent Technologies
12:40 Lattice-based Unsupervised MLLR for Speaker Adaptation
Mukund Padmanabhan, George Saon, Geoffrey Zweig, IBM
13:00-14:30 Lunch

Tuesday Afternoon
8 Acoustic Modeling, 14:30-16:10
14:30 Hidden-Articulator Markov Models for Speech Recognition 
Matt Richardson, Jeff Bilmes, Chris Diorio, Univ. of Washington
14:50 Performance of Mandarin Connected Digit Recognizer with Word Duration Modeling
Gang Peng, Bo Zhang, William S-Y. Wang, Dept. of Electronic Engineering, City Univ. of Hong Kong
15:10 Rate-of-Speech Modeling for Large Vocabulary Conversational Speech Recognition 
Jing Zheng, Horacio Franco, Andreas Stolcke, SRI International
15:30 Lightly Supervised Acoustic Model Training 
Lori Lamel, Jean-Luc Gauvain, Gilles Adda, LIMSI-CNRS
15:50 Cross-Language Transfer of Multilingual Phoneme Models 
A.K. Kienappel, D. Geller, R. Bippus, Philips Research Laboratories
16:10-16:40 Coffee break
9 Noise Robustness II, 16:40-18:00
16:40 Noise Adaptation of HMMs Using Neural Networks
Sadaoki Furui, Daisuke Itoh, Tokyo Institute of Technology, Dept. of Computer Science
17:00 Amplitude Modulation Maps for Robust Speech Recognition 
G.F. Meyer, B.A. Edmonds, D. Yang, W.A. Ainsworth, Keeele Univ.
17:20 From Multi-Band Full Combination to Multi-Stream Full Combination Processing in Robust ASR 
Astrid Hagen, Andrew Morris, Hervé Bourlard, Dalle Molle Institute for Perceptual Artificial Intelligence (IDIAP)
17:40 The AURORA Experimental Framework For The Performance Evaluation of Speech Recognition Systems Under Noisy Conditions
H.-G. Hirsch, Ericsson Eurolab and D. Pearce, Motorola Labs
20:00 Conference dinner

Wednesday Morning
10  Keynote, 9:40-10:30
9:40 Towards Super-Human Speech Recognition (slides,ppt)
Mukund Padmanabhan, Michael Picheny, IBM
10:30-11:00 Coffee break
11 Error Analysis, Confidence Measures and Metadata, 11:00-13:00
11:00 Linguistic Dissection of Switchboard-Corpus Automatic Speech Recognition Systems
Steven Greenberg, Shawn Chang, ICSI
11:20 Optimizing confidence measure based on HMM acoustical rescoring 
Delphine Charlet, France Telecom R&D
11:40 Prosodically Motivated Features for Confidence Measures
Silke Goronzy, Krzysztof Marasek, Andreas Haag, Ralf Kompe, Sony Int'l (Europe)
12:00 Recognition Confidence Scoring for Use in Speech Understanding Systems 
Timothy J. Hazen, Theresa Burianek, Joseph Polifroni, Stephanie Seneff, MIT Laboratory for Computer Science
12.20 Model Selection Criteria for Acoustic Segmentation
Mauro Cettolo, Marcello Federico, ITC-IRST, Centro per la Ricerca Scientifica e Tecnologica
12.40 Sentence Boundary Detection in Broadcast Speech Transcripts
Yoshihiko Gotoh, Steve Renals, Univ. of Sheffield
13:00-14:30 Lunch

Wednesday Afternoon
12 Keynote, 14:30-15:20
14:30 Surfing the Voice Web: Issues in the design of a voice browser 
Michael H. Cohen, Nuance Communications
13 New National and International Speech Projects and Future Prospects, 15:20-19:30
15:20 Human Language Technologies in the European Research Programmes, current state and future perspectives 
Mats Ljungqvist, EC
15:40 Breaking the Local Optima Paradigm: DARPA Speech Research Initiatives in Multi-Modal and other Technologies 
Jim Bass, DARPA
16:00 A Japanese national project on spontaneous speech corpus and processing technology 
Sadaoki Furui, Kikuo Maekawa, Hitoshi Isahara
16:20 ASR for Automatic Directory Assistance: the SMADA Project 
Lou Boves, Denis Jouvet, Juergen Sienel, Renato de Mori, Frederic Bechet, Luciano Fissore, Pietro Laface
16:40-17:10 Coffee break
17:10-18:00 Short presentations from projects (not included in proceedings)
     ALERT, Gerhard Rigoll, Duisburg Univ. 
     NESPOLE!, Mauro Cettolo, IRST 
     Meeting transcription corpus (ppt) John Garofolo, NIST 
     CATCH2004, Ponani Gopalakrishnan, IBM
     SPEECON, Hans-Gunter Hirsch, Ericsson
18:00-19:30 Outstanding challenges and future directions (Coretex project, RWTH, CUED, IRST, LIMSI)

©LIMSI-CNRS, Orsay-France, 2000
Last modified: Monday,12-November-01 01:17:55 CET (stats)