Keynotes
Optical Compressive Multidimensional Sensing
The theory of Compressive Sensing (CS) offers a sensing framework aimed at capturing data in a more efficient manner compared with traditional sensing approaches. The CS theory has found natural applications in optics. Utilizing CS for optical multidimensional sensing is particularly appealing since the data is typically in high dimension and compressible, and because there is no hardware, which can capture directly more than two dimensions. In this talk, we will overview CS techniques developed for various multidimensional sensing modalities, such as 3D coherent and incoherent sensing, spectral imaging, and spectral light field imaging.
Adrian Stern received his Ph.D. degree in Electrical Engineering from Ben-Gurion University of the Negev in 2003. Currently he is full Professor at the Electro-Optics Engineering Department at Ben-Gurion University in Israel, where he serves as Department Head. He has held visiting scholar positions at MIT and UConn. His current research interests include compressive imaging and optical sensing, 3D imaging, computational imaging, spectral imaging, phase-space optics.
Dr. Stern is a Fellow of OSA and SPIE, and a Member of IEEE. He has served as an Associate Editor for several journals and is the Editor of the book “Optical compressive imaging” that appeared in 2016.
Inverse problems in light field imaging
The task in inverse problems is to reconstruct a signal from observations that are subject to a known (or inferred) corruption process. Examples of inverse problems are denoising, de-blurring, super-resolution, and reconstruction from a sparse set of measurements. To address inverse problems in 2D imaging, that are typically ill-posed, many low-dimensional models besides sparsity have become popular to incorporate various flavors of prior knowledge: (structured) sparsity, low-rank, manifold structure, graph structure. Solving inverse problems indeed requires a good understanding of the structure of the image space. Machine learning techniques allowing us to learn the latent space in which the images reside, from a set of examples, are now conquering this field of research.
While there are now very mature techniques for solving inverse problems in 2D imaging, the processing of light fields, due to the large volume of high dimensional data that they represent and their specific structure, require solutions beyond a straightforward application of 2D image processing methods. This talk will review techniques exploiting various data priors for solving inverse problems in light field imaging with a focus on denoising, super-resolution, inpainting.
Christine Guillemot is currently Director of Research at INRIA (Institut National de Recherche en Informatique et Automatique) in France. She holds a PhD degree from ENST (Ecole Nationale Supérieure des Telecommunications) Paris (1992). From 1985 to 1997, she has been with France Telecom working in the areas of image and video compression for multimedia and digital television. From 1990 to mid 1991, she has worked as visiting scientist at Bellcore Bell Communication research) in the USA. Her research interests are signal and image processing, and in particular 2D and 3D image and video processing for various problems (compression, super-resolution, inpainting, classification). She has co-authored 25 patents, has published 80 journal publications and 190 publications in peer reviewed international conferences. She received an ERC advanced grant for a project on computational imaging (2016-2021).
She has served in both the IEEE MMSP technical committee (2005-2008), and the IEEE IVMSP technical committee (2013-2016). She has been Associate Editor for IEEE Trans. on Image Processing (from 2000 to 2003 and from 2014 to 2016), for IEEE Trans. on Circuits and Systems for Video Technology (2004-2006), IEEE Trans. on Signal Processing (2007-2009). for the Eurasip Journal on Image Communication (2010-2016), and member of the IEEE Journal on Selected Topics in Signal Processing (2013-2016). She is currently senior area editor for IEEE Trans. on Image Processing and senior member of the steering committee of IEEE Trans. on Multimedia. She is IEEE Fellow since January 2013.
Radon Transform for Practical Light Field Imaging
Todor Georgiev is Principal Scientist at Adobe. Previously, he has held Principal Engineer and Senior Research Scientist positions at Qualcomm and Adobe. He is well known of his works on plenoptic and light field camera systems and corresponding processing methods.