Tag: sparse reconstruction
Local maxima estimation of gradients to speed up a convex optimization algorithm
Contact: Adrian Jarret (LCAV) Synopsis: Discover the Frank-Wolfe algorithm for convex optimization with sparse priors, contribute to the development of a fast polyatomic variant, find local maxima on a 2D image, simulate image reconstruction for radio astronomy. Level: Bachelor – Master (semester project) Sections: IC Description: The LASSO problem [1] is a convex optimization problem (…)
Signal reconstruction using sparsity-promoting penalty over a redundant dictionary of wavelets
Contact: Adrian Jarret (LCAV) Synopsis: Comparison of analysis and synthesis formulations, solution of analysis formulation using cycle spinning with different wavelet basis, solution of synthesis formulation with adjoint of penalty operator and proximal gradient method, extension to weighted penalty. Level: Master (5-6 months) Sections: IC Description: Wavelet bases are known to produce sparse representations of (…)