Go to the list of all software

Signal Reconstruction on GPUs

Description:

This software implements two algorithms on graphical processing units (GPUs), using CUDA platform of NVIDIA. [1] In compressed sensing, we implemented a simple version of the SpaRSA algorithm described by Wright, Nowak, and Figueiredo, using randomly chosen rows of discrete cosine transform (DCT) matrix as the sensing matrix. [2] In image denoising/deblurring, we implemented the PDHG algorithm of Zhu and Chen, which solves the Rudin-Osher-Fatemi formulation with total variation regularization.

Link:

http://pages.cs.wisc.edu/~swright/GPUreconstruction/

Software File:

Authors:

Lee, Sangkyun