Guillaume Tochon

Guillaume Tochon is an Assistant Professor at EPITA since 2016 and an Invited Researcher at the Institut de Mécanique Céleste et de Calcul des Éphémérides (IMCCE, Observatoire de Paris) since 2022. He holds a M.Sc. and Ph.D. in electrical engineering and signal/image processing from Grenoble Institute of Technology and Université Grenoble Alpes, with postdoctoral experience at Carnegie Institution for Science and UCLA. His research focuses on mathematical morphology, statistical signal processing, and machine/deep learning, applied to space imagery (e.g., astrometry) and satellite data analysis. He leads projects on morphological neural networks, Sentinel-2 time series dynamics, and meteor/satellite detection using 1-second exposure imagery.

• Education:
• M.Sc. in Electrical Engineering, Grenoble Institute of Technology (2012)
• Ph.D. in Signal and Image Processing, Université Grenoble Alpes (2015)
• Visiting Scholar, UCLA Department of Mathematics (2013)

Research Interests: Combines mathematical morphology with machine learning to address challenges in remote sensing and space imagery. Key areas include hierarchical representations, noise analysis, and applications in astrometry and hyperspectral unmixing. Current projects involve neural Koopman operators for forecasting, Sentinel-2 time-series assimilation, and meteor detection in night-sky imagery.

Teaching: At EPITA, he teaches Mathematics for Signal Processing, Data Compression, Image Processing, Convex Optimization, Machine Learning, and Constrained Optimization across undergraduate and graduate levels.

Collaborations: Works with researchers at IMCCE (e.g., V. Lainey, J. Vaubaillon), IMT-Atlantique, and Grenoble-INP. Supervises PhD students in topics such as noise estimation, spectral dynamics, and Cassini image analysis.

Labs/Teams: Active in EPITA Research Laboratory's Image Processing and Pattern Recognition team and IMCCE's PEGASE team.