Research activities

Research interests

• Statistical signal and image processing.
• Inverse problems in signal and image processing.
• Myopic and blind source separation.
• Sparse representations.
• Anomaly detection.
• Model selection
• Manifold learning.

Tools and methods

• Bayesian inference, hierarchical models.
• Latent variable models, Markov random fields.
• Bayesian nonparametric models, Gaussian processes, marked point processes.
• Markov chain Monte Carlo (MCMC) algorithms: Metropolis-Hastings, Hamiltonian Monte Carlo, reversible jump MCMC.

Applications

• Hyperspectral imagery. Endmember extraction. Spectral unmixing. Nonlinear mixture and anomaly detection.
• Multispectral Lidar. Surface detection et spectral analysis of semi-transparent targets
• Low illumination imaging

Collaborations (past and current)

• José M. BIOUCAS DIAS ( Instituto Superior Tecnico, Portugal).
• José C. M. BERMUDEZ (Federal Univ. Santa Catarina, Florianopolis, Brazil).
• Gerald S. BULLER (Heriot-Watt University, Edinburgh, UK).
• Nicolas DOBIGEON (University of Toulouse, IRIT/INP-ENSEEIHT, Toulouse, France).
• Alfred HERO (University of Michigan, Ann Arbor, Michigan)
• Steve McLAUGHLIN (Heriot-Watt University, Edinburgh, UK).
• Miles Padgett (University of Glasgow, Glasgow, UK).
• Marcelo PEREYRA (Heriot-Watt University, Edinburgh, UK).
• Cédric RICHARD (University of Nice, Lagrange Lab., Nice, France).
• Jean-Yves TOURNERET (University of Toulouse, IRIT/INP-ENSEEIHT, Toulouse, France).
• Andrew WALLACE (Heriot-Watt University, Edinburgh, UK).