EuromatrixPlus, Bringing machine translation for European languages to use, 2009-2012. This European FP7 project joins together the main laboratories in Europe in the field of automatic translation: University of Edinburgh (Scotland), University of Dublin (Ireland), University of Prague (Czech Republic), Université du Maine (France), DFKI (Germany), FBK (Italy), and LUCY (Germany). The project approaches many aspects to advance research in statistical machine translation. The LIUM participates mainly in sub-project "advanced learning methods for machine translation".
GALE, Global Autonomous Language Exploitation, 2009-2011. This DARPA international project involves 9 partners under the coordination of IBM. The project started in June 2006. The LIUM was invited by IBM to join their consortium in September 2009, following our excellent results during the 2009 NIST translation evaluation campaign. The project aims at transcription, translation and analysis of broadcasts in Arabic and Chinese for the American military intelligence.
MateCAT, Machine Translation Enhanced Computer Assisted Translation, 2011-2014.
BOLT, Broad Operational Language Translation, 2011-2014. This DARPA international project involves 9 partners under the coordination of IBM.
EUMSSI, Event Understanding through Multimodal Social Stream Interpretation, 2013-2016. Projet FP7. The main objective of EUMSSI is developing technologies for identifying and aggregating data presented as unstructured information in sources of very different nature (video, image, audio, speech, text and social context), including both online (e.g., YouTube) and traditional media (e.g. audiovisual repositories), and for dealing with information of very different degrees of granularity.
JOKER, Towards social and affective relations with a robot, 2013-2017. projet Europe - CHIST-ERA. This project will build and develop a generic intelligent user interface providing a multimodal dialogue system with social communication skills including humor, empathy, compassion, charm, and other informal socially-oriented behavior.
SODA, 2010-2014. This ANR project (call for projects REPERE) involves the LIUM (coordinator), and the IDIAP. REPERE is an evaluation campaign of people recognition technologies for audio-visual French TV shows. The competitors will have to propose systems based on various information sources (audio and video data, closed-captions, internet images) to determine who speaks, who appears in the images, which people names are pronounced or appear on the screen, and to which people those names correspond.
COSMAT, Collaborative Service of Automatic Translation for Scientific Texts, 2009-2012. This ANR project involves the LIUM (coordinator), SYSTRAN, and the INRIA. The applicative context is the open, multi-disciplinary archive HAL (scientific research papers, theses, …). The work focuses on hybrid translation systems, systems dedicated to specific field, and the use of corrections made by users.
ASH, 2009-2012. This ANR (French National Research Agency) project involves the LIUM (coordinator), the IRISA (Rennes) and the LIA (Avignon). The theme of the project is collaboration between transcription systems.
PORT-MEDIA, 2009-2012. This ANR project, led by the LIA (Avignon), with collaboration from the LIG (Grenoble), ELRA (Paris), the LORIA (Nancy) and the LIUM, focuses on the issues of robustness and portability of multilingual spoken language understanding systems. The LIUM provides the transcription system and studies the interdependence between the transcription system and the dialog system.
DEPART, 2009-2013. This project, funded by the French region Pays de la Loire, involves the LINA (coordinator, Nantes), IRCCyN (Nantes) and the LIUM. The objective of the project is to build a regional network of competences that associates the analysis of the audio signal and of handwriting with automatic language processing.
EPAC, 2006-2010. This ANR project is coordinated by the LIUM and involves the LIA (Avignon), the IRIT (Toulouse) and the LI (Tours). It focuses on extraction and processing of conversational speech from audio documents.
VERA, 2012-2016. The VERA project aims at developing tools for diagnostic, localization, and measurements of automatic transcription errors. This project is based on a consortium of first-rate academic actors in this field. The objective is to study the errors in detail (at the perceptive, acoustico-phonetics, lexical, and syntactic levels) in order to yield a precise diagnosis of possible lacks of the current classical models on certain classes of linguistic phenomena.
PEA TRAD, 2011-2014. Le projet PEA-TRAD porte sur la traduction pour l’aide à l’analyse documen- taire. Le projet est financé par la DGA et a pour partenaires Cassidian/EADS, Systran, le LIG, Université d’Edimbourg, le LNE, ELDA et le LIUM.
Voxolab is a spinoff of the LIUM in charge of the commercial development of speaker and speech recognition tools.