Research Interests

My personal research interests are Automatic Speech Recognition (ASR) and Statistical Machine Translation (SMT).
Currently, I am an associate professor in the Language and Speech Technology at the LIUM laboratory.


Master's degree of Mathematics and Computer Science, Artificial Intelligence and the Web Option AIW
at the Joseph Fourier University of Grenoble UJF and the "Institut National Polytechnique de Grenoble" INPG

Master thesis

Subject : Spoken Language Translation of Arabic

Spoken language translation of Arabic language has been widely studied recently in different projects (DARPA TRANSTAC, GALE) or evaluation campaigns (IWSLT, NIST). Most of the time, the rich morphology of Arabic language is seen as a problem that must be addressed, especially when dealing with sparse data. It has been shown that pre-processing Arabic data using a morphological segmenter is useful to improve machine translation results. we proposed to use simultaneously multiple segmentations for machine translation of Arabic. The core idea is to keep the ambiguity of the Arabic segmentation at the system input (using confusion networks).

Laboratory : Laboratory of Informatics of Grenoble LIG.

Team : Study Group for Machine Translation and Automated Processing of Languages and Speech GETALP.

Supervisors : Laurent Besacier & Hervé Blanchon.

Publications :
1) LIG approach for IWSLT09 : Using Multiple Morphological Segmenters for Spoken Language Translation of Arabic.
2) Traduction automatique de la parole arabe/anglais par segmentations multiples. RJCP, Avignon(France), 17-19 Nov 2009

Report :


Subject : Harnessing of Heteregenous ASR Systems

Laboratory : LIUM

Team : LST

Supervisors : Paul Delèglise, Yannick Estève and Georges Linarès

Harnessing of Heteregenous ASR Systems (ASH Project)

The ASH project is a research project started on october 2009 and funded by the French National Research Agency (ANR) for 36 months. The ASH project proposes a new approach to combine several ASR systems. This approach consists in making these ASR systems exchange information during the decoding process, on the fly, whereas classical approach consists in only combining final outputs. This implies the design of a specific framework, and allows the use of different techniques.

In the ASH project, the landmark-driven technique will be part which can be integrated in this new framework.

Technical skills

Machine Translation : Moses statistical machine translation system

Arabic pre-processing tools : Buckwalter Morphological analyser, Mada, SVM

Speech recognizers : CMU sphinx toolkit, Speeral toolkit and RWTH RASR toolkit

Machine learning techniques : n-gram models, Hiden Markov Models, Clustering

Systems : Linux, Macos, Windows.

Laboratoire d'Informatique de l'Université du Maine (LIUM)
Institut d'Informatique Claude Chappe
Université du Maine, Avenue Laënnec