Sites Inria

Version française

PANAMA Research team

Parcimonie et Nouveaux Algorithmes pour le Signal et la Modélisation Audio

Team presentation

At the interface between audio modeling and mathematical signal processing, the global objective of the PANAMA project-team is to develop mathematically founded and algorithmically efficient techniques to model, acquire and process high-dimensional signals, with a strong emphasis on acoustic data.

Research themes

Research activities are structured around the following axes: - sparse models and representations - robust acoustic scene analysis - large-scale audio content processing and self-organization

International and industrial relations

The project team has regular industrial collaborations through CIFRE Ph.D. contracts with Orange Labs and Technicolor, and through bilateral and collaborative projects with Audionamix, Studio Maia, Sonic Emotion, and Canon. These collaborations, together with the team's contributions to the SISEC/SASEC evaluation campaigns, bring relevant feedback from field deployment. Long term international academic collaborations have been established with EPFL, University of Edinburgh, Queen Mary, University of London, and Technion, in the context of international collaborative projects.

Keywords: Sparse modeling; audio and music signal processing; source separation; machine learning; inverse problems