Stéphane HUET

Assistant professor (ATER)
IRISA / University of Rennes 1
PhD advisors : Guillaume Gravier and Pascale Sébillot

Contact

IRISA
Campus de Beaulieu
35042 RENNES Cedex
France

tel: +33 2 99 84 75 45
fax: +33 2 99 84 71 71
e-mail: Stephane.Huet [at] irisa.fr

Secretary: Loïc Lesage, +33 2 99 84 74 37, Loic.Lesage [at] irisa.fr

Research interests

Natural language processing
Automatic speech recognition
Indexing

Current research activities

I'm studying the application of the methods of Natural Language Processing (NLP) to Automatic Speech Recognition (ASR) systems. As NLP is generally done on witten corpora, one of my first goals is analyzing its behavior on oral corpora. Another objective is using information extracted thanks to NLP to inmprove the rates of recognition of an ASR system.
The data I'm working on are essentially automatically transcribed news text.
During my PhD, I more specifically studied the use of morpho-syntactic information and thematic adaptation to improve ASR. I worked under the supervision of Pascale Sébillot and Guillaume Gravier into the TEXMEX team at IRISA. My thesis also proceeded in collaboration with the METISS team at IRISA and the "Institut national de l'audiovisuel".

Abstract of my thesis

A way to improve outputs produced by automatic speech recognition (ASR) systems is to integrate additional linguistic knowledge. Our research in this field focuses on two aspects: morpho-syntactic information and thematic adaptation.

In the first part, we propose a new mode of integration of parts of speech in a post-processing stage of speech decoding. To do this, we tag N-best sentence hypothesis lists with a morpho-syntactic tagger built to take into account the specificities of transcriptions. We reorder these lists by modifying the score computed by an ASR system at the sentence level to include morpho-syntactic information. Experiments done on French-speaking broadcast news (ESTER corpus) exhibit a significant improvement of the word error rate. Besides, we establish the contribution of morpho-syntactic information to improve posterior based confidence measures.

In the second more exploratory part, we are interested in thematically adapting the language model (LM) of an ASR system. We propose a scheme that enables us to specialize speech decoding in an unsupervised way. We first segment the studied document into thematically homogeneous sections. To this end, we develop a new probabilistic framework to integrate different modalities (lexical cohesion, acoustic clues, and linguistic markers) and show its relevance to improve segmentation. We then build adaptation corpora retrieved from the Web by using an innovative procedure. We finally modify the LM with these specific corpora and show that, on thematic sections that are manually selected, this method significantly improves the LM, even if the increase of the word error rate is slight.

Publications

Complete list

Teaching

2004:
- Java at the INSA of Rennes (practical works: 14h)
2004-2005:
- HTML and Javascript at the University of Rennes 1 (practical works: 20h)
- Operating systems and imperative programming with Java at the University of Rennes 1 (tutorials: 20h, practical works: 40h)
2005-2006:
- Imperative programming with Eiffel at the University of Rennes 1 (tutorials: 24h, practical works: 24h)
- Functional programming with Scheme at the University of Rennes 1 (tutorials: 20h, practical works: 20h)
2006-2007:
- Functional programming with Scheme at the University of Rennes 1 (practical works: 20h)
- Spreadsheet use at the University of Rennes 1 (practical works: 4h)
- Networks and protocols (HDLC, Ethernet, IP, TCP, UDP, sockets...) (tutorials: 24h, practical works:24h)
2007-2008:
- Statistical hypothesis test at the University of Rennes 1 (pratical works: 16h)
- Networks and protocols (HDLC, Ethernet, IP, TCP, UDP, sockets...) at the University of Rennes 1 (tutorials: 24h, practical works:24h)
- Operating systems (8086, memory adress, file system, input/output) at the University of Rennes 1 (lecture: 4h, tutorials: 16h)
- Object-oriented programming with Java at the University of Rennes 1 (tutorials: 8h, pratical works: 14h)

Short biography

I received my engineering diploma of the Institut National des Sciences Appliquées (INSA) of Rennes in 2004 and my master thesis at the University of Rennes1 (IFSIC) the same year. I defended my PhD in computer science on the 11^th of december 2007.

I did the practical period of my master thesis in the SIMBAD team at IRISA, where I studied myoelectric signals to order an artificial hand.