Algorithms for audio inpainting based on probabilistic nonnegative matrix factorization

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26220%2F22%3APU146638" target="_blank" >RIV/00216305:26220/22:PU146638 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.sciencedirect.com/science/article/pii/S0165168422004443" target="_blank" >https://www.sciencedirect.com/science/article/pii/S0165168422004443</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1016/j.sigpro.2022.108905" target="_blank" >10.1016/j.sigpro.2022.108905</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Algorithms for audio inpainting based on probabilistic nonnegative matrix factorization

Popis výsledku v původním jazyce

Audio inpainting, i.e., the task of restoring missing or occluded audio signal samples, usually relies on sparse representations or autoregressive modeling. In this paper, we propose to structure the spectrogram with nonnegative matrix factorization (NMF) in a probabilistic framework. First, we treat the missing samples as latent variables, and derive two expectation–maximization algorithms for estimating the parameters of the model, depending on whether we formulate the problem in the time- or time-frequency domain. Then, we treat the missing samples as parameters, and we address this novel problem by deriving an alternating minimization scheme. We assess the potential of these algorithms for the task of restoring short- to middle-length gaps in music signals. Experiments reveal great convergence properties of the proposed methods, as well as competitive performance when compared to state-of-the-art audio inpainting techniques.

Název v anglickém jazyce

Algorithms for audio inpainting based on probabilistic nonnegative matrix factorization

Popis výsledku anglicky

Audio inpainting, i.e., the task of restoring missing or occluded audio signal samples, usually relies on sparse representations or autoregressive modeling. In this paper, we propose to structure the spectrogram with nonnegative matrix factorization (NMF) in a probabilistic framework. First, we treat the missing samples as latent variables, and derive two expectation–maximization algorithms for estimating the parameters of the model, depending on whether we formulate the problem in the time- or time-frequency domain. Then, we treat the missing samples as parameters, and we address this novel problem by deriving an alternating minimization scheme. We assess the potential of these algorithms for the task of restoring short- to middle-length gaps in music signals. Experiments reveal great convergence properties of the proposed methods, as well as competitive performance when compared to state-of-the-art audio inpainting techniques.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

20203 - Telecommunications

Projekt

<a href="/cs/project/GA20-29009S" target="_blank" >GA20-29009S: Restaurace degradovaných audiosignálů založená na sluchové percepci</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2022

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

SIGNAL PROCESSING

ISSN

1872-7557

e-ISSN

—

Svazek periodika

neuveden

Číslo periodika v rámci svazku

206

Stát vydavatele periodika

NL - Nizozemsko

Počet stran výsledku

10

Strana od-do

1-10

Kód UT WoS článku

000922000100001

EID výsledku v databázi Scopus

2-s2.0-85146050781