Speech Translation System – Low-resource Languages to High-resource Languages

 

Supervisors: Aghilas Sini (LIUM), Jane Wottawa (LIUM)
Hosting lab: LIUM (Laboratoire d’Informatique de l’Université du Mans)
Place: Le Mans Université
Beginning of internship : March 2024
Contacts: Aghilas Sini and Jane Wottawa, (firstname.name@univ-lemans.fr)

Application: Send a CV, a covering letter relevant to the proposed subject, your grades for the last two years of study and the possibility of attaching letters of recommendation to all the supervisors, before November 15, 2024
 
 

Research context :
We are looking for a student to work on the TV2M-E project, which aims to study automatic speech translation systems. The focus is on language pairs that are poorly or under-resourced (Kabyle, Kurdish, Tamashaq) and hight-resourced (French, English, German). The candidate will analyse the balance of power between the language pairs as they emerge from the systems. Various aspects will be addressed :

• Lexicon and fixed expressions,
• Phonetics and sound proximity between source and target,
• Syntax and paraphrasing in the target language.

The project is also looking at phono-stylistic issues in a broad sense, such as the translation of interjections (haha, hé), vocal identity in translation and other distinctive factors between source and target..

Profile required :
The candidate will be a student in the humanities or language sciences and will work on the project during the M1 and M2 periods on a paid basis. He or she will also write a Master’s thesis on the work carried out as part of the project at the LIUM, LST team. The person recruited will be interested in aspects of speech and machine translation as well as stylistic analysis of speech.
Previous knowledge of machine translation is not required. On the other hand, knowledge of one of the above-mentioned poorly or under-resourced languages would be a plus.

 

Working environment :
L’encadrement de la recherche et du mémoire seront assurés par Aghilas Sini, enseignant-chercheur en informatique et porteur du projet, et Jane Wottawa, enseignante-chercheuse en Linguistique. Les heures de travail seront assurées au laboratoire du LIUM (bâtiment IC2) et avec un ordinateur fourni par le projet.

 

 

Key-words : automatic translation, under-resourced languages, analyses linguistiques et phono-stylistiques