Short and squeezed: accelerating the computation of antisparse representations with safe squeezing
Abstract
Antisparse coding aims at spreading the information uniformly over representation coefficients and can be expressed as the solution of an ∞-norm regularized problem. In this paper, we propose a new methodology, coined "safe squeezing", accelerating the computation of antisparse representations. The idea consists in identifying saturated entries of the solution via simple tests and compact their contribution to achieve some form of dimensionality reduction. Numerical experiments show that the proposed approach leads to significant computational gain.
Origin : Files produced by the author(s)