Publication detail

Structured Sparsity for Audio Inpainting

MACH, V.

Original Title

Structured Sparsity for Audio Inpainting

Type

conference paper

Language

English

Original Abstract

Completing missing or distorted samples is a typical task of audio restoration engineers. State-of-the-art interpolation techniques utilized for this purpose are compromised today by novel methods based on sparse representations of signals. Preliminary results of sparse solutions are promising. In this paper, the topic of sparse solutions of underdetermined linear systems is described in a theoretical way. Further, the comparison of interpolation and sparse solutions based on l1 -minimization is presented.

Keywords

Audio, Inpainting, Interpolation, Structured, Sparsity, l1 -minimization.

Authors

MACH, V.

RIV year

2014

Released

24. 4. 2014

Location

Brno

ISBN

978-80-214-4924-4

Book

Proceedings of the 20th Conference STUDENT EEICT 2014

Pages from

41

Pages to

45

Pages count

5

BibTex

@inproceedings{BUT107261,
  author="Václav {Mach}",
  title="Structured Sparsity for Audio Inpainting",
  booktitle="Proceedings of the 20th Conference STUDENT EEICT 2014",
  year="2014",
  pages="41--45",
  address="Brno",
  isbn="978-80-214-4924-4"
}