EURASIP logo

EURASIP Summer School on Remote Sensing and Microscopy Image Processing

29 July - 2 August 2024, Veszprem, Hungary

Outline

The main focus of the Summer School is on image and signal processing methods applied for remote sensing and microscopy imaging. While these two topics seem far away from each other, in reality there is a considerable overlap between the problems and techniques applied to these fields. The reason is that in both cases the ratio of the size of objects that we are observing and the distance from where images are taken is extremely small. Many of the invited lecturers were actively involved in the research of both domains, sometimes adopting the same methodology for both problems.
The invited lecturers will cover the following selected topics in this field:

  • probabilistic graphical models
  • CNNs and deep learning
  • semantic segmentation
  • object detection (e.g. single cell extraction, road network detection,...)
  • biodiversity tracking
  • radar remote sensing
  • aspects of polarization
  • shape models
  • multimodal data fusion
  • marked point process
  • synthetic aperture radar (SAR)
  • ship and ship wake detection
  • single cell-based large-scale microscopy
  • machine learning methods to identify various cellular phenotypes

SSRM2024 Speakers and Lectures

Sparse Distributions and Inverse Problems in SAR Imaging of the Sea Surface

Alin Achim, University of Bristol, UK

Abstract
In space imaging, enhanced image quality is key to the detection and characterisation of difficult and transient targets. For example, accurate evaluation of the sea surface conditions can help with the detection and characterisation of ship wakes. These provide key information for tracking (illegal) vessels and are also useful in classifying the characteristics of the wake generating vessel. Until recently, one of the main factors hampering research into sea surface modelling was the lack of data of sufficiently high resolution (pixels need to be typically smaller than few meters) and accuracy. Remote-sensing technologies have however shown remarkable progress in recent years and the availability of remotely sensed data of the Earth and sea surface is continuously growing. Several European missions (e.g., the Italian COSMO/SkyMed, the German TerraSAR-X, or the British NovaSAR) have developed a new generation of satellites exploiting synthetic aperture radar (SAR) to provide spatial resolutions previously unavailable from space-borne remote sensing.

In this talk, we will first concentrate on the statistical characterisation of high-resolution SAR images of the sea surface, and we will show how these can be best described using heavy-tailed distributions based on alpha-stable models. We will subsequently describe techniques for enhancing the quality of SAR images, by solving inverse problems that directly exploit the sparsity of data in specific transform domains and show their benefits on ship target detection. Then, we will introduce a new method for detecting linear features, which are characteristic of typical ship wakes. We address this as a sparse estimation problem using both convex and non-convex optimisation techniques based on the Radon transform and sparse regularisation. This breaks into subproblems which are solved using proximal algorithms. Finally, we will show automatic target recognition (ATR) results on real ship images using an AI system trained on artificial images created using an in-house developed simulator.

Keywords: Inverse problems, synthetic aperture radar (SAR), ship and ship wake detection, ATR.

Bio
Alin Achim (Senior Member, IEEE) received the B.Sc. and M.Sc. degrees in electrical engineering from the "Politehnica" University of Bucharest, Bucharest, Romania, in 1995 and 1996, respectively, and the Ph.D. degree in biomedical engineering from the University of Patras, Patras, Greece, in 2003. He then obtained an European Research Consortium for Informatics and Mathematics (ERCIM) Postdoctoral Fellowship which he spent with the Institute of Information Science and Technologies (ISTI-CNR), Pisa, Italy, and with the French National Institute for Research in Computer Science and Control (INRIA), Sophia Antipolis, France. In October 2004, he joined the Department of Electrical and Electronic Engineering, University of Bristol, Bristol, U.K., as a Lecturer, he became a Senior Lecturer (Associate Professor) in 2010, and a Reader in biomedical image computing in 2015. Since August 2018, he holds the Chair in computational imaging at the University of Bristol, and from 2019 to 2020, he was a Leverhulme Trust Research Fellow with the Laboratoire I3S, Université Cote d'Azur, Nice, France. He has coauthored over 200 scientific publications, including more than 50 journal articles. His research interests include statistical signal, image, and video processing, with particular emphasis on the use of sparse distributions within sparse domains and with applications in both biomedical imaging and remote sensing. Dr. Achim was/is an Elected Member of the Bio Imaging and Signal Processing Technical Committee of the IEEE Signal Processing Society, an Affiliated Member (invited) of the IEEE Signal Processing Society's Signal Processing Theory and Methods Technical Committee, and a member of the IEEE Geoscience and Remote Sensing Society's Image Analysis and Data Fusion Technical Committee. He is Senior Area Editor of the IEEE Transactions on Image Processing, an Associate Editor of the IEEE Transactions on Computational Imaging, and an Editorial Board Member of Remote Sensing (MDPI).

Urban scene perception and environment model synthesis from multisensorial spatial data

Csaba Benedek, HUN-REN SZTAKI, Hungary

Abstract
In the past decade we have witnessed an explosion of new technologies for acquisition and understanding of environmental information. 3D vision and perception systems of self-driving vehicles, mobile robots, and drones can be used for - apart from safe navigation - real time mapping of the environment, detecting and analyzing static and dynamic scene elements. On the other hand, new generation geo-information systems (GIS) store extremely detailed 3D maps about the cities, consisting of dense 3D point clouds, registered camera images and semantic metadata. 

In this lecture, I present new techniques to facilitate the joint exploitation of the measurements of mobile online sensing platforms, and offline environmental data obtained by various 3D mapping technologies in urban environment. The introduced methods comprise a Lidar based real time and accurate self-localization and change detection approach, a fully automatic, online camera- Lidar calibration technique, a novel deep neural network-based change detection approach, which can robustly extract changes between sparse and weakly registered points, and machine learning based realistic virtual augmentation of initially low resolution and incomplete camera images, 2.5D range images and 3D point cloud measurements.

Bio
Prof. Csaba Benedek is a deputy director of the HUN-REN Institute for Computer Science and Control (SZTAKI), and a scientific advisor (D.Sc.) with the Machine Perception Research Laboratory of HUN-REN SZTAKI, where he is the head of the Research Group on Geo-Information Computing. He also works as a full professor with the Faculty of Information Technology and Bionics of the Péter Pázmány Catholic University. Between 2008 and 2009 he was a postdoctoral researcher at INRIA Sophia-Antipolis, working in the Ariana Project Team with Prof. Josiane Zerubia. Csaba Benedek is a vice chairman of the John von Neumann Computer Society, a past president of the Hungarian Image Processing and Pattern Recognition Society (Képaf), and the current Hungarian Governing Board Member of the International Association for Pattern Recognition (IAPR). He is a Senior Member of IEEE, and an Associate Editor of Digital Signal Processing (Elsevier) journal. He received various awards including the Bolyai plaquette from the Hungarian Academy of Sciences (2019) and the Michelberger Master Award from the Hungarian Academy of Engineering (2020). He has been the manager of various national and international research projects in the recent years. His research interests include Bayesian image and point cloud segmentation, object extraction, change detection, machine learning applications and GIS data analysis. His book titled "Multi-level Bayesian Models for Environment Perception" was published by Springer in 2022.

Webpage of the research group: http://mplab.sztaki.hu/geocomp/demos.html

Aspects of Polarization in Earth Observation Using Radar Remote Sensing

Avik Bhattacharya, IIT Bombay, India

Abstract
Polarization plays a pivotal role in radar remote sensing, shaping our understanding of Earth's surface properties by analyzing electromagnetic wave interactions. This lecture explores vital aspects of polarization in radar remote sensing, from the basics to the advanced realm of polarimetric synthetic aperture radar (PolSAR). By examining the polarization states, information content, and target characterization techniques, we delve into the wealth of insights polarization provides about surface structures, orientations, and material compositions. The lecture also highlights the applications of polarimetric radar in various fields, such as agriculture, forestry, and environmental monitoring. Despite its complexities, polarization emerges as a powerful tool, enabling researchers to discern intricate details and unravel Earth's mysteries from above.

Bio
Avik Bhattacharya (Senior Member, IEEE) received the integrated M.Sc. degree in mathematics from the Indian Institute of Technology, Kharagpur, India, in 2000, and the Ph.D. degree in remote sensing image processing and analysis from Télécom ParisTech, Paris, France, and the Ariana Research Group, Institut National de Recherche en Informatique et en Automatique (INRIA), Sophia Antipolis, Nice, France, in 2007.,He is a Professor at the Centre of Studies in Resources Engineering, Indian Institute of Technology Bombay (CSRE, IITB), Mumbai, India. Before joining IITB, he was a Canadian Government Research Fellow at the Canadian Centre for Remote Sensing (CCRS), Ottawa, ON, Canada. He received the Natural Sciences and Engineering Research Council of Canada Visiting Scientist Fellowship at the Canadian National Laboratories from 2008 to 2011. His research interests include SAR polarimetry, statistical analysis of polarimetric SAR images, and applications of radar remote sensing in agriculture, cryosphere, urban, and planetary studies. Dr. Bhattacharya was the Associate Editor and Editor-in-Chief of IEEE Geoscience and Remote Sensing Letters (GRSL) from 2019 to 2023. He has been the Guest Editor of the special issue on Applied Earth Observations and Remote Sensing, India in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing (J-STARS), since 2017. He was one of the Guest Editors of the special stream on Advanced Statistical Techniques in SAR Image Processing and Analysis in IEEE Geoscience and Remote Sensing Letters, in 2018. He is the Founding Chairperson of the IEEE Geoscience and Remote Sensing Society (GRSS) Chapter of the Bombay Section. He leads the Microwave Remote Sensing Laboratory (www.mrslab.in) at CSRE, IITB.

Sparse optimization for inverse problems in imaging. Application in super-resolution microscopy

Laure Blanc-Féraud,  DR CNRS, laboratoire I3S, France

Abstract
The course will introduce inverse problems in imaging with examples in microscopy imaging, a.e. deconvolution and super-resolution. Regularization of inverse problems is presented in the variational approach, describing L1-norm, L0-pseudo norm, TV norm and related algorithms in the scope of convex/non convex, smooth/non smooth optimization. Practical examples will be described for biological fluorescence microscopy imaging.

Objects detection in biological image processing: from marked point processes to deep learning 

Xavier Descombes, INRIA, Sophia Antipolis, France

Abstract
In this talk I will review the main principles of marked point processes for analyzing images. I will illustrate the approach on numerous applications from remote sensing to biomedical images. I will then state the pros and cons of this approach and the deep learning one. A comparison of the two approaches will be given for detecting nuclei in histpathological images of kidney. Finally I will give some ideas to combine deep learning with marked point process modelling.

Bio
Xavier Descombes received the bachelor's degree in telecommunications from the Ecole Nationale Superieure des Telecommunications de Paris (ENST) in 1989, the master of science degree in mathematics from the University of Paris VI in 1990, the PhD degree in signal and image processing from the ENST in 1993, and the "habilitation" degree from the University of Nice-Sophia Antipolis in 2004. He has been a postdoctoral researcher at ENST in 1994, at the Katholieke Universiteit Leuven in 1995, and at INRIA in 1996, and a visiting scientist at the Max Planck Institute of Leipzig in 1997. Positions:

1994 : Research Assistant at Telecom Paris
1995 : Free Research Fellow at KUL (Leuven, Belgium)
1996 : Post-doctoral Fellow at INRIA Sophia Antipolis (PASTIS team) 
1997 : Visiting Scientist at MPI Leipzig, Germany 
1997-2011 : Permanent researcher at INRIA Sophia Antipolis (Ariana team)
2011-present : Head of Morpheme team at INRIA Sophia Antipolis 

Life beyond the pixels: deep learning methods in cancer and virus research

Peter Horvath, BRC, Szeged, Hungary

Abstract
In this talk I will give an overview of the computational steps in the analysis of single cell-based large-scale microscopy experiments. First, I will present a novel microscopic image correction method designed to eliminate illumination and uneven background effects which, left uncorrected, corrupt intensity-based measurements. New single-cell image segmentation methods will be presented using differential geometry, energy minimization and deep learning methods (www.nucleAIzer.org). I will discuss the Advanced Cell Classifier (ACC) (www.cellclassifier.org), a machine learning software tool capable of identifying cellular phenotypes based on features extracted from the image. It provides an interface for a user to efficiently train machine learning methods to predict various phenotypes. For cases where discrete cell-based decisions are not suitable, we propose a method to use multi-parametric regression to analyze continuous biological phenomena. To improve the learning speed and accuracy, we propose an active learning scheme that selects the most informative cell samples. I will present our recently developed single-cell isolation methods, based on laser-microcapturing and patch clamping, that utilize the selection and extraction of specific cell(s) using the above machine learning models.

Bio
Director, Group Leader, Principal Investigator of Biological Image Analysis and Machine Learning Group (BIOMAG), BRC, Szeged, Hungary. Education and positions

2023 - present Senior Researcher - Helmholtz Munich, AI4Health Institute, Germany
2018 - present Director, Institute of Biochemistry Biological Research Centre
2017 - 2023 Director, FIMM High-Content Analysis Facility, FIMM-HCA HELMI
2014 - 2023 Finnish Distinguished Professor Fellowship, FIMM, Helsinki
2013 - present Biological Image Analysis and Machine Learning Group (BIOMAG), BRC, Szeged
2013 Senior Scientist, Kutay Group, Institute of Biochemistry, ETH, Zurich
2007 - 2012 Lecturer, Light Microscopy and Screening Centre, ETH, Zurich
2004 - 2007 Ph.D. INRIA, Sophia Antipolis, France and University of Szeged

Bayesian statistical shape analysis (and related ideas)

Ian H. Jermyn, University of Durham, UK

Abstract
Shape has long been a subject of study, from Aristotle, through D'Arcy Thompson, to more recent statistical and machine learning work. The statistical analysis of shape proper began in the 1960's with finite sets of points, moving on to curves and surfaces and beyond, and more recently to 'geometric deep learning' and related formulations. The earlier ideas, rooted as they are in geometry and group theory, are still relevant to the more recent work, and the earlier methods continue to be used in many domains, notably where there is a paucity of data or a need for scientific understanding. This talk begins at the very beginning and presents statistical shape analysis as the study of a series of increasingly complex objects using increasingly complex models. The statistical approach is Bayesian, and this provides the excuse for a brief foray into the foundations of that subject.

Bio
Ian H. Jermyn is currently a professor in statistics with the Department of Mathematical Sciences, Durham University. His research interests centre around geometrical questions in statistics, on methodologies for the representation and statistical modelling of geometric structure, and on the geometrical underpinnings essential to the success of statistical shape analysis. Methodologies include Riemannian geometry, applied to the description of shape spaces, and random fields with long-range and higher-order interactions, used in particular to model objects with complex and varying topologies. He has also worked substantially on applications of these methodologies, in particular to image processing and computer vision.

Semantic segmentation of remote sensing images through the integration of deep learning and probabilistic graphical models

Gabriele Moser, University of Genoa, Italy

Abstract
In the framework of land cover mapping applications, the semantic segmentation of remote sensing images plays a primary role. On one hand, methods based on deep learning have currently become the dominant approaches to this task. However, their accuracy is often affected by the quantity and quality of the ground truth for training, which may often be scarce in land cover mapping applications. On the other hand, probabilistic graphical models represent a family of powerful stochastic models for image analysis and structured output learning, thanks to their capability to model complex dependencies within Bayesian processing tasks. In this seminar, the potential of the integration of the deep learning and probabilistic graphical model approaches is discussed in the context of the semantic segmentation of remote sensing imagery. The rationale of their joint use and combination will be presented methodologically, and examples of experimental results will be shown in applications to land cover mapping from aerial very high-resolution optical images and multimission radar images.

Bio
Gabriele Moser (Senior Member, IEEE) received the Laurea (M.Sc. equivalent) degree in telecommunications engineering and the Ph.D. degree in space sciences and engineering from the University of Genoa, Genoa, Italy, in 2001 and 2005, respectively. Since 2001, he has been cooperated with the Image Processing and Pattern Recognition for Remote Sensing (IPRS) Laboratory, University of Genoa, where he has been an Associate Professor of telecommunications since 2014. He has also been the Head of the RIOS Laboratory, Savona, Italy, since 2013. He has authored or coauthored more than 140 articles in international scientific journals and conference proceedings and edited books. His research interests include pattern recognition and image processing methodologies for remote sensing and energy applications.

Biolmage Analysis: Methods, Tools and Applications 

Jean-Christophe Olivo-Marin, Pasteur Institute, France

Abstract
We present recent topics developed in the laboratory related to spatial statistics [1-2] and mechanobiology [3-4]. The quantitative analysis of molecule interactions in bioimaging is key for understanding the molecular orchestration of cellular processes and is generally achieved through the study of the spatial colocalization between different populations of molecules. Most colocalization methods are based on pixel overlap between the previously denoised signal that is emitted from two (or more) different fluorescent labels and use a global image correlation such as Pearson's or Manders' coefficients. These data, however, cannot be linked to physical parameters such as the real percentage of colocalizing molecules or the average colocalization distance. In addition, randomly distributed molecules can partially overlap, and it is hard to measure the statistical significance of the computed correlation indices. We will present a novel statistical method (SODA) [1], to analyse molecule colocalization that is based on the automatic detection of molecule fluorescent spots, followed by their representation as Point Processes and the statistical analysis of their spatial distribution. We will illustrate the method through examples in TIRF and 3D-STORM microscopy. We will also present Generalized SODA (GSODA) [2] a derived method to measure statistical colocalization in the paradigm of level set analysis.

In a second part, we will present methods and tools that allow quantitative analyses of motion and/or shape changes of biological objects from image time series [3]. We will illustrate the retrieval of mechanical quantities that govern cell mechanics from image sequence data of living cells, and show results in tracking, measuring and modelling the motility and shape dynamics of amiba [4].

Bibliography

1. Lagache, T., Grassart, A., Dallongeville, S., Faklaris, O., Sauvonnet, N., Dufour, A., Danglot, L., Olivo-Marin, J.-C. (2018) Mapping molecular assemblies with fluorescence microscopy and object-based spatial statistics, Nature Communications, 9,1, pp. 698-711. doi: 10.1038/s41467-018-03053-x.

2. Mukherjee, S., Gonzalez-Gomez, C., Danglot, L., Lagache, T. and and Olivo-Marin, J.-C. (2020) Generalizing the Statistical Analysis of Objects' Spatial Coupling in Bioimaging, IEEE Signal Processing Letters, 27, 6, pp. 1085-9. doi: 10.1109/LSP.2020.3003821

3. Boquet-Pujadas, A., Lecomte, T., Manich, M., Thibeaux, R., Labruyere, E., Guillén, N., Olivo-Marin*,  .J - C., and Dufour*, A.C. (2017) BioFlow: a non-invasive, image-based method to measure speed, pressure and forces inside living cells. Scientific Reports, 7, 1, 9178. doi: 10.1038/s41598-017-09240-y.

4. Boquet-Pujadas, A., Feaugas, T., Petracchini, A., Grassart, A., Mary, H., Manich, M., Gobaa, S., Olivo-Marin, J.-C., Sauvonnet, N., and Labruyere, E. (2022) 4D live imaging and computational modeling of a functional gut-on-a-chip evaluate how peristalsis facilitates enteric pathogen invasion. Science Advances, 8, 42. eabo5767. doi: 10.1126/sciadv.abo5767

Remote sensing in Earth observation for biodiversity tracking and road throughput optimization

Tamas Sziranyi, SZTAKI, Budapest, Hungary

Abstract
Satellite image data of multispectral channels can help us to estimate the terrestrial soil, water content and  plant conditions. These informations are very helpful when  soil moisture should be analysed for offroad vehicle throughput or tracking the living conditions of plants. On the one hand, this is important for monitoring the water balance and biodiversity, and on the other hand, in the event of possible flooding, for estimating the passability of roads and changes in vegetation.

Bio
Tamas Sziranyi (Senior Member, IEEE) received the Ph.D. and D.Sci. degrees from the Hungarian Academy of Sciences, Budapest, in 1991 and 2001, respectively. He was appointed as a Full Professor in 2001 at Pannon University, Veszprém, Hungary, and in 2004, at Pázmány Péter Catholic University, Budapest. He has been a Research Scientist at the Institute for Computer Science and Control (SZTAKI) since 1992, where he has been leading the Machine Perception Research Laboratory since 2006. He is currently a Full Professor at the Budapest University of Technology and Economics. He has participated in several prestigious international (ESA, EDA, FP6, FP7, and OTKA) projects with his research laboratory. He has more than 310 publications, including 60 in major scientific journals and several international patents. His research activities include machine perception, pattern recognition, texture and motion segmentation, Markov random fields and stochastic optimization, remote sensing, surveillance, intelligent networked sensor systems, graph-based clustering, and digital film restoration. He was the Founder and the Past President (1997-2002) of the Hungarian Image Processing and Pattern Recognition Society. Since 2008, he has been a fellow of the International Association for Pattern Recognition (IAPR) and the Hungarian Academy of Engineering. He has been a member of Hungarian Academy of Sciences since 2022. He was honored by the Master Professor Award in 2001, the ProScientia (Veszprem) Award in 2011, and the Officers Cross by the President of Hungary in 2018. He was an Associate Editor of IEEE Transactions on Image Processing (2003-2009). He has been an Associate Editor and now Area Editor of Digital Signal Processing (Elsevier) since 2012.

Marked Point Process Models in Image Processing. Application to Remote Sensing

Josiane Zerubia, INRIA, Sophia Antipolis, France

Abstract
Stochastic methods are now widespread in image analysis. They have proved to be powerful tools to solve inverse problems such as image classification or restoration. Since the mid-nineties, many research works have extended the initial pixel based approach to the concept of object in order to deal with shape detection problems. In particular, stochastic models have shown good potentialities in extracting simple shapes. Generally, configurations of parametric functions are sampled from probability distributions defined in a configuration space, Markov Chain Monte Carlo (MCMC) being one of the most popular families of samplers. In various application domains, from line detection to 3D reconstruction, the MCMC samplers are efficient for object extraction in large configuration spaces from any type of probability distribution. Models based on marked point processes are among the most efficient stochastic approaches and have lead to convincing experimental results in various shape detection applications (such as extraction of line segments, rectangles, circles, ellipses, .). The marked point processes exploit random variables whose realizations are configurations of geometrical objects. After specifying a probability distribution measuring the quality of each object configuration, the maximum density estimator is searched for by MCMC techniques coupled with a stochastic relaxation. Such processes are especially adapted to the description of complex spatial interactions between the objects. Various examples on road network detection, crown tree extraction, flamingo counting, boat detection or building reconstruction will be given during the talk. Then, learning point process models for vehicles detection using Convolutional Neural Networks (CNNs) in satellite images will be described. Finally, results on both high-resolution satellite and drone images will be presented.

Bio
Josiane Zerubia has been a permanent research scientist at INRIA since1989 and Director of Research since July 1995 (DR Exceptional Class since February 2023; DR 1st Class from 2002 to 2022). She was head of the PASTIS remote sensing laboratory (INRIA Sophia-Antipolis) from mid-1995 to 1997 and of the ARIANA research group (INRIA/CNRS/University of Nice), which worked on inverse problems in remote sensing and biological imaging, from 1998 to 2011. From 2012 to 2016, she was head of AYIN research group (INRIA-SAM) dedicated to models of spatio-temporal structure for high-resolution image processing with a focus on remote sensing and skincare imaging. She is head of AYANA exploratory research group since 2020. AYANA is an interdisciplinary project using knowledge in stochastic modeling, image processing, artificial intelligence, remote sensing, and embedded electronics/computing. 

She was professor (PR1) at SUPAERO (ISAE) in Toulouse from 1999 to2020. She received a Doctor Honoris Causa degree from the University of Szeged in Hungary in 2020, and 3 times the Excellence Award from University of Nice (now UCA) in 2020, 2019, and 2016.

She supervised or co-supervised 63 Master students, 37 PhDs, and 27 post-docs. She was external examinator for PhD degrees at Purdue Univ. (West-Lafayette, USA), Heriot Watt Univ. (Edinburgh, GB), Univ. of Iceland (Reyljavik, Iceland), University of Lisbon (Portugal), Univ. of Manouba (Tunis, Tunisia), Sup Telecom (Tunis, Tunisia), University of Rabat (Morocco), and for more than 30 PhDs in France including one at the University of the French West Indies.

Before that, she was with the Signal and Image Processing Institute of the University of Southern California (USC) in Los-Angeles as a postdoc (1988-1989). She also worked as a researcher for the LASSY (University of Nice/CNRS) from 1984 to 1988 and in the Research Laboratory of Hewlett Packard in France and in Palo-Alto (CA) from 1982 to 1984. She received the MSc degree from the Department of Electrical Engineering at ENSIEG, INP Grenoble, France in 1981, the Doctor of Engineering degree, her PhD and her 'Habilitation', in 1986, 1988, and 1994 respectively, all from the University of Nice, France.

She is a Fellow of the IEEE (2003-), the EURASIP (2019-) and the IAPR (2020-), and IEEE SP Society Distinguished Lecturer (2016-2017). 

She was a member of the IEEE IMDSP TC (SP Society) from 1997 to 2003, of the IEEE BISP TC (SP Society) from 2004 to 2012 and of the IVMSP TC (SP Society) from 2008 to 2013. She was associate editor of IEEE Trans. on IP from 1998 to 2002, area editor of IEEE Trans. on IP from 2003 to 2006, guest co-editor of a special issue of IEEE Trans. on PAMI in 2003, member of the editorial board of IJCV from 2004 to March 2013 and member-at-large of the Board of Governors of the IEEE SP Society from 2002 to 2004. She was also associate editor of the on-line resource "Earthzine" (IEEE CEO and GEOSS) from 2006 to mid-2018. She was a member of the editorial board of the French Society for Photogrammetry and Remote Sensing (SFPT) from 1998 to 2020, and member-at-large of the Board of Governors of the SFPT from 2014 to 2020. She was a member of the IEEE Signal Processing Magazine Senior editorial board from September 2018 to January 2022. She was member-at-large of the Awards Board of the IEEE SP Society from 2020 to 2022. Finally, she was a member of the Best Paper Award Committee for EURASIP JIVP in 2021 and also member of the IAPR Fellow committee in 2021 and 2022.

She has been a member of the editorial board of the Foundation and Trends in Signal Processing since 2007 and of the IEEE WISP Committee since 2024.

She was co-chair of two workshops on Energy Minimization Methods in Computer Vision and Pattern Recognition (EMMCVPR'01, Sophia Antipolis, France, and EMMCVPR'03, Lisbon, Portugal), co-chair of a workshop on Image Processing and Related Mathematical Fields (IPRM'02, Moscow, Russia), technical program chair of a workshop on Photogrammetry and Remote Sensing for Urban Areas (Marne-la-Vallée, France, 2003), co-chair of the special sessions at IEEE ICASSP 2006 (Toulouse, France) and IEEE ISBI2008 (Paris, France), publicity chair of IEEE ICIP 2011 (Brussels, Belgium), tutorial co-chair of IEEE ICIP 2014 (Paris, France), general co-chair of the EarthVision workshop at IEEE CVPR 2015 (Boston, USA) and a member of the organizing committee and plenary talk co-chair of IEEE-EURASIP EUSIPCO 2015 (Nice, France). She also organized and chaired an international workshop on Stochastic Geometry and Big Data at Sophia Antipolis, France, in 2015. She was part of the organizing committees of the EarthVision workshop (co-chair) at IEEE CVPR2017 (Honolulu, USA) and GRETSI 2017 symposium (Juan-les-Pins, France). She was scientific advisor and co-organizer of ISPRS 2020 (virtual), 2021 (virtual), and 2022 congress (Nice, France) and technical co-chair of IEEE-EURASIP EUSIPCO 2021 (virtual, Dublin, Ireland). 

Her main research interest is in image processing using probabilistic models. She also works on parameter estimation, statistical learning, optimization techniques, and artificial intelligence.
Last modified: Thu Mar 28 21:55:46 CET 2024