Publication detail

Reweighted l1 minimization for audio inpainting

MOKRÝ, O. RAJMIC, P.

Original Title

Reweighted l1 minimization for audio inpainting

English Title

Reweighted l1 minimization for audio inpainting

Language

en

Original Abstract

Reweighted l1 minimization has been successfully applied to the task of audio declipping by Weinstein and Wakin (2011). We adapt the reweighting to the audio inpainting problem, for both the synthesis and the analysis models. It is shown that the reweighting provides a significant improvement in terms of SNR, especially in the analysis model.

English abstract

Reweighted l1 minimization has been successfully applied to the task of audio declipping by Weinstein and Wakin (2011). We adapt the reweighting to the audio inpainting problem, for both the synthesis and the analysis models. It is shown that the reweighting provides a significant improvement in terms of SNR, especially in the analysis model.

Keywords

Inpainting; Sparse; Cosparse; Synthesis; Analysis

Released

01.07.2019

Location

Toulouse, France

Pages count

2

URL

Documents

BibTex


@inproceedings{BUT157345,
  author="Ondřej {Mokrý} and Pavel {Rajmic}",
  title="Reweighted l1 minimization for audio inpainting",
  annote="Reweighted l1 minimization has been successfully applied to the task of audio declipping by Weinstein and Wakin (2011). We adapt the reweighting to the audio inpainting problem, for both the synthesis and the analysis models. It is shown that the reweighting provides a significant improvement in terms of SNR, especially in the analysis model.",
  chapter="157345",
  howpublished="online",
  year="2019",
  month="july"
}