Axe “Science des données”, 17 décembre 2020

Map Unavailable

Date(s) - 17/12/2020
14 h 00 - 16 h 00


Title : Physics aware explainable AI for Earth Observation
Author : Mihai Datcu
Abstract : Generally imaging sensors generate an isomorphic representation of the observed scene. This is not the case for Earth Observation. A satellite image is a doppelgänger of the scattered field, an indirect signature of the imaged object. The lecture presents a concept based on a hybrid approach. The EO images are decomposed based on the a priori knowledge of physical models to make evidence of the image formation process and scene scattering properties. Further unsupervised methods extract the innate data models, generating categories of canonical signatures. However, since the observed scenes have high complexity, not all parameters can be estimated. To cope with this complexity, supervised DNN architectures are used to learn the models of the images and their transformations. Finally, in an embedding process images semantically labeled are combined with explicitly specified physical scattering classes or domain knowledge.  The methods are exemplified with multispectral and Synthetic Aperture Radar (SAR) observations.

Mihai Datcu received the M.S. and Ph.D. degrees in Electronics and Telecommunications from the University Politechnica Bucharest UPB, Romania, in 1978 and 1986. In 1999 he received the title Habilitation à diriger des recherches in Computer Science from University Louis Pasteur, Strasbourg, France. Currently he is Senior Scientist and Data Intelligence and Knowledge Discovery research group leader with the Remote Sensing Technology Institute (IMF) of the German Aerospace Center (DLR), Oberpfaffenhofen, and Professor with the Department of Applied Electronics and Information Engineering, Faculty of Electronics, Telecommunications and Information Technology, UPB. From 1992 to 2002 he had a longer Invited Professor assignment with the Swiss Federal Institute of Technology, ETH Zurich. From 2005 to 2013 he has been Professor holder of the DLR-CNES Chair at ParisTech, Paris Institute of Technology, Telecom Paris. His interests are in Data Science, Machine Learning and Artificial Intelligence, and Computational Imaging for space applications. He is involved in Big Data from Space European, ESA, NASA and national research programs and projects. He is a member of the ESA Big Data from Space Working Group. He received in 2006 the Best Paper Award, IEEE Geoscience and Remote Sensing Society Prize, in 2008 the National Order of Merit with the rank of Knight, for outstanding international research results, awarded by the President of Romania, and in 1987 the Romanian Academy Prize Traian Vuia for the development of SAADI image analysis system and activity in image processing. He is IEEE Fellow. He has been holder of a 2017 Blaise Pascal International Chair of Excellence at CEDRIC, CNAM.

Title : An Overview on Music Information Retrieval
Author : Francesco Foscarin
Abstract : Music information retrieval (MIR) is the interdisciplinary science of retrieving information from music. While this field of research has been explored for decades, the popularity of the topic has massively grown over the last year and is now widely applied in multiple commercial applications. This presentation starts with some examples of popular MIR tasks and then moves to an overview on the most common data representations formats for music. Finally, we present the MIR research problems that we have targeted in past research at CNAM and the topics we are currently exploring.