Open-source codes associated with the publications I co-authored.

Inverse Problems in Biomedical Imaging

GlobalBioIm Library
A Matlab library for solving inverse problems.
[Code] [Paper] [Website]
Structured Illumination Microscopy (SIM)
  • FlexSIM A flexible SIM reconstruction method capable to handle difficult data prone to reconstruction artifacts. [Code] [Paper]
  • Single-slice SIM Reconstruction algorithm enabling the estimation of the out-of-focus signal and based on an inner-loop-free ADMM. [Code] [Paper]
Single Molecule Localization Microscopy (SMLM)
  • Sliding Frank-Wolfe: Grid-less method that minimizes the BLASSO. [Code] [Paper]
  • SMLM-CEL0: Grid-based method that relies on the minimization of the Continuous Exact L0 functional. [Code] [Paper]
  • Continuous FRC: Computes the Fourier-ring correlation (FRC) in the continuous domain for single molecule localization microscopy. [Code] [Paper]
Optical Diffraction Tomography (ODT)
Reconstruction method based on the exact Lippmann-Schwinger forward model and the FISTA algorithm. [Code] [Paper Opt. Exp.] [Paper TCI]
Multi-Angle Total Internal Reflection Fluorescence Microscopy (MA-TIRFM)
Reconstruction algorithm that performs jointly volume reconstruction, deconvolution, and background estimation. [Code] [Paper Sci. Rep.] [Paper ISBI]
Single-Particle Cryo Electron Microscopy (SP Cryo-EM)
Joint angular refinement and reconstruction algorithm for single-particle Cryo-EM.
[code] [Paper]
Image Denoising
Automatic Tuning of Denoising Algorithms Parameters without Ground Truth.
[code] [Paper]

Sparse Optimization

Safe screening with sphere refinement for sparse optimization
Safe screening approach that exploits local strong concavity properties of the dual function to accelerate the convergence of iterative solvers. [Code] [Paper JMLR] [Paper SPL]
Sliding Frank-Wolfe algorithm
Grid-less method that minimizes the BLASSO. [Code] [Paper]
L2 - CEL0 Minimization
Examples of the use of the FBS and IRL1 algorithms to solve the CEL0 penalized least squares problem. [Code] [Paper SIOPT] [Paper SIIMS]

Non-negative Matrix Factorization (NMF)

Transform Learning NMF
  • Illustrations of our paper on the identifiability of TL-NMF. [Code] [Paper]
  • Comparisons between TL-NMF and Joint Diagonalization + NMF. [Code] [Paper]

Emmanuel Soubies - 2014