Avatar

Stamatis Lefkimmiatis

Senior Research Scientist

Q Bio Inc

Biography

I am a Senior Research Scientist in the Radiomics R&D team with Q Bio, where I am responsible for developing algorithms and methods that improve Medical Imaging. From 2016 to 2019, I was an Assistant Professor at Skolkovo Institute of Science and Technology (Skoltech), Moscow and director of the Computational Imaging Group (CIG). I have also held Postdoctoral Research Associate positions with the Biomedical Imaging Group at EPFL and the Department of Applied Mathematics at UCLA.

My research interests lie in the areas of computational imaging, machine & deep learning, computer vision and large-scale optimization. My current focus is on inverse imaging problems, including, denoising, deconvolution, super-resolution, demosaicking, compressive sensing, etc., with applications in digital photography, biomicroscopy and medical imaging.

For more information about me, see my CV or contact me.

Interests

  • Computational Imaging
  • Computer Vision
  • Machine & Deep Learning
  • Large-Scale Optimization

Education

  • PhD in Electrical & Computer Engineering, 2009

    National Technical University of Athens (NTUA)

  • Diploma/M.Eng. in Computer Engineering & Informatics, 2004

    University of Patras, Greece

Professional Experience

 
 
 
 
 

Senior Research Scientist

Q Bio Inc.

Feb 2019 – Present San Francisco, CA, USA
Responsible for leading the R&D efforts of the Radiomics team in the areas of parallel and quantitative MRI reconstruction.
 
 
 
 
 

Assistant Professor

Skolkovo Institute of Science and Technology (Skoltech)

May 2016 – Mar 2019 Moscow, Russia
Principal investigator of the Computational Imaging Group (GIG). Supervised 3 PhD candidates and 9 Msc students. Lectures on fundamental topics in signal and image processing. Research in computational imaging, deep-learning, and low-level computer vision with applications in digital photography and biomedical imaging.
 
 
 
 
 

Postdoctoral Research Fellow

University of California, Los Angeles (UCLA)

May 2014 – Apr 2016 Los Angeles, CA, USA
Member of the Computational and Applied Mathematics Group, working with Prof. Stanley Osher. Research in non-local variation methods and convex optimization techniques for large-scale inverse imaging problems.
 
 
 
 
 

Postdoctoral Research Associate

École polytechnique fédérale de Lausanne (EPFL)

Jul 2010 – Apr 2014 Lausanne, Switzerland
Member of the Biomedical Imaging Group, working with Prof. Michael Unser. Research in the design of novel regularization methods and large-scale convex optimization techniques with emphasis on biomedical and computer-vision applications.

Honors & Awards

2017 Best Paper Award

Best paper award for the paper :

  • S. Lefkimmiatis, A. Bourquard and M. Unser, Hessian-Based Norm Regularization for Image Restoration With Biomedical Applications, IEEE Transactions on Image Processing, Volume 21, No. 3, March 2012

Advanced Postdoc Mobility Fellowship

Thomaidio Award

Award for the paper :

  • S. Lefkimmiatis, P. Maragos and G. Papandreou, Bayesian inference on multiscale models for Poisson intensity estimation: Applications to photon-limited image denoising, IEEE Trans. Image Process., vol. 18, no. 8, pp. 1724–1741, August 2009.

PhD scholarship

PhD scholarship based on academic achievements

TEE Award

Award to top ranking students in the School of Computer Engineering Informatics

IKY Award

Award to top ranking students in the School of Computer Engineering & Informatics for the academic year 2002-2003

Selected Publications

Full list @ Google Scholar

Quickly discover relevant content by filtering publications.

Maxwell Parallel Imaging

Purpose: To develop a general framework for Parallel Imaging (PI) with the use of Maxwell regularization for the estimation of the …

Microscopy Image Restoration with Deep Wiener-Kolmogorov filters

Microscopy is a powerful visualization tool in biology, enabling the study of cells, tissues, and the fundamental biological processes; …

Iterative Joint Image Demosaicking and Denoising Using a Residual Denoising Network

Modern digital cameras rely on the sequential execution of separate image processing steps to produce realistic images. The first two …

Iterative Residual CNNs for Burst Photography Applications

Modern inexpensive imaging sensors suffer from inherent hardware constraints which often result in captured images of poor quality. …

Universal Denoising Networks : A Novel CNN Architecture for Image Denoising

We design a novel network architecture for learning discriminative image models that are employed to efficiently tackle the problem of …

News