Section: Contracts and Grants with Industry
ARC INRIA RAPSODIS
Participant : Guillaume Gravier.
Duration: 2 years, starting in February 2008. Partners: Metiss , Parole , Talaris project-teams, CEA-LIST/LIC2M.
The Concerted Research Action RAPSODIS (Syntactic and Semantic Information-Based Automated Speech Recognition) aims at improving automatic speech recognition (ASR) by integrating linguistic information. Based on former work by S. Huet concerning the incorporation of morpho-syntactic knowledge in a post-processing stage of the transcription, we experiment, together with our partners, the deep insertion of automatically obtained semantic relations (especially paradigmatic ones) and syntactic knowledge within an ASR system.
In 2008, work has been mostly dedicated to the study of possible integration modes—reordering of n-best hypothesis lists is currently privileged—, to investigations about the impact of transcription errors on syntactic parsing—a correlation between the length of chunks in hypothesis and errors has been established—, to the acquisition of semantic relations from the Web, and to a major refactoring of our ASR system, using a larger amount of training data, in order to enable effective integration of linguistic information.
QUAERO CTC and Corpus Projects
Main academic partners : IRCAM, IRIT, LIMSI, TelecomParisTech, Univ. Karlsruhe, CLIPS/Imag.
Quaero is a European research and development program with the goal of developing multimedia and multilingual indexing and management tools for professional and general public applications (such as search engines). The project was approved by The European Commission on 11 March 2008.
This program is supported by OSEO. The consortium is led by Thomson. Other companies involved in the consortium are: France Télécom, Exalead, Bertin Technologies, Jouve, Grass Valley GmbH, Vecsys, LTU Technologies, Siemens A.G. and Synapse Développement. Many public research institutes are also involved, including LIMSI-CNRS, INRIA, IRCAM, RWTH Aachen, University of Karlsruhe, IRIT, Clips/Imag, Telecom ParisTech, INRA, as well as other public organisations such as INA, BNF, LIPN and DGA.
METISS is involved in two technological domains : audio processing and music information retrieval (WP6). The research activities (CTC project) are focused on improving audio and music analysis, segmentation and description algorithms in terms of efficiency, robustness and scalability. Some effort is also dedicated on corpus design, collection and annotation (Corpus Project).
METISS also takes part to research and corpus activities in multimodal processing (WP10), in close collaboration with the Texmex project-team.