Nonnegative sparse representation based on the determinant measure | IEEE Conference Publication | IEEE Xplore

Nonnegative sparse representation based on the determinant measure


Abstract:

Recently, sparse representations via an overcomplete dictionary has become a major field of researches in signal processing. This paper focuses on sparse representation o...Show More

Abstract:

Recently, sparse representations via an overcomplete dictionary has become a major field of researches in signal processing. This paper focuses on sparse representation of nonnegative signals since signals and corresponding dictionary have nonnegativity limitations in some applications, e.g., multispectral data analysis. We present a novel sparsity measure based on a kind of determinant of matrix. Unlike the conventional sparsity measures, the proposed measure is differentiable and easy to optimize. Based on this measure, a new sparse model is derived, and an iterative sparseness minimization approach is proposed to solve this model. In the approach, the nonnegative sparse representation problem can be cast into row-to-row optimizations with respect to the sparse coefficient matrix, and then the quadratic programming (QP) is used to optimize each row. Numerical experiments on recovery of the sparse coefficient and synthesis dictionary show the effectiveness of the proposed algorithm.
Date of Conference: 16-18 October 2016
Date Added to IEEE Xplore: 02 March 2017
ISBN Information:
Electronic ISSN: 2165-3577
Conference Location: Beijing, China

References

References is not available for this document.