From the site:
TV-Normed Pursuit is a new sparse recovery paradigm where efficient total-variation (TV) constrained algorithms from combinatorial and convex optimization interface for interpretable signal reconstruction. The algorithm is presented in "Hard thresholding with norm constraints", Technical Report, by Anastasios Kyrillidis, Gilles Puy and, Volkan Cevher.In this wiki page, we provide information about TV-Normed Pursuit algorithm and present experiments on real image data. We illustrate that TV-Normed Pursuit can significantly enhance the performance of both combinatorial methods and convex solvers in total-variation (TV) constrained sparse recovery.
The paper: HARD THRESHOLDING WITH NORM CONSTRAINTS by Anastasios Kyrillidis, Gilles Puy, and Volkan Cevher. The abstract reads:
We introduce a new sparse recovery paradigm, called NORMED PURSUITS, where efﬁcient algorithms from combinatorial and convex optimization interface for interpretable and model-based solutions. Synthetic and real data experiments illustrate that NORMED PURSUITS can signiﬁcantly enhance the performance of both hard thresholding methods and convex solvers in sparse recoveryThe wiki and code are here.
Liked this entry ? subscribe to Nuit Blanche's feed, there's more where that came from. You can also subscribe to Nuit Blanche by Email, explore the Big Picture in Compressive Sensing or the Matrix Factorization Jungle and join the conversations on compressive sensing, advanced matrix factorization and calibration issues on Linkedin.