New Methods for l2-l0 Minimization and their Applications to 2D Single-Molecule Localization Microscopy

We present in this paper a biconvex reformulation of a ℓ
₂
-ℓ
₀
problem composed of a least-square data term plus a sparsity term introduced as a constraint or a penalization. Minimization algorithms are derived and compared with the state of the art in ℓ
₂
-ℓ
₀
minimization by relaxation or deep learning. Application results are shown on Single-Molecule Localization Microscopy.