Griffin-lim like phase recovery via alternating direction method of multipliers

Yoshiki Masuyama, Kohei Yatabe, Yasuhiro Oikawa

Research output: Contribution to journalArticle

Abstract

Recovering a signal from its amplitude spectrogram, or phase recovery, exhibits many applications in acoustic signal processing. When only an amplitude spectrogram is available and no explicit information is given for the phases, the Griffin-Lim algorithm (GLA) is one of the most utilized methods for phase recovery. However, GLA often requires many iterations and results in low perceptual quality in some cases. In this letter, we propose two novel algorithms based on GLA and the alternating direction method of multipliers (ADMM) for better recovery with fewer iteration. Some interpretation of the existing methods and their relation to the proposed method are also provided. Evaluations are performed with both objective measure and subjective test.

LanguageEnglish
Article number8552369
Pages184-188
Number of pages5
JournalIEEE Signal Processing Letters
Volume26
Issue number1
DOIs
Publication statusPublished - 2019 Jan 1

Fingerprint

Method of multipliers
Alternating Direction Method
Recovery
Spectrogram
Acoustic signal processing
Iteration
Signal Processing
Acoustics
Evaluation

Keywords

  • Non-convex optimization
  • phaseless spectrogram inversion
  • short-time Fourier transform (STFT)
  • spectrogram consistency
  • STFT-based speech synthesis

ASJC Scopus subject areas

  • Signal Processing
  • Electrical and Electronic Engineering
  • Applied Mathematics

Cite this

Griffin-lim like phase recovery via alternating direction method of multipliers. / Masuyama, Yoshiki; Yatabe, Kohei; Oikawa, Yasuhiro.

In: IEEE Signal Processing Letters, Vol. 26, No. 1, 8552369, 01.01.2019, p. 184-188.

Research output: Contribution to journalArticle

@article{5d0ce61ee6a743cb93ae1d3df1b11270,
title = "Griffin-lim like phase recovery via alternating direction method of multipliers",
abstract = "Recovering a signal from its amplitude spectrogram, or phase recovery, exhibits many applications in acoustic signal processing. When only an amplitude spectrogram is available and no explicit information is given for the phases, the Griffin-Lim algorithm (GLA) is one of the most utilized methods for phase recovery. However, GLA often requires many iterations and results in low perceptual quality in some cases. In this letter, we propose two novel algorithms based on GLA and the alternating direction method of multipliers (ADMM) for better recovery with fewer iteration. Some interpretation of the existing methods and their relation to the proposed method are also provided. Evaluations are performed with both objective measure and subjective test.",
keywords = "Non-convex optimization, phaseless spectrogram inversion, short-time Fourier transform (STFT), spectrogram consistency, STFT-based speech synthesis",
author = "Yoshiki Masuyama and Kohei Yatabe and Yasuhiro Oikawa",
year = "2019",
month = "1",
day = "1",
doi = "10.1109/LSP.2018.2884026",
language = "English",
volume = "26",
pages = "184--188",
journal = "IEEE Signal Processing Letters",
issn = "1070-9908",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "1",

}

TY - JOUR

T1 - Griffin-lim like phase recovery via alternating direction method of multipliers

AU - Masuyama, Yoshiki

AU - Yatabe, Kohei

AU - Oikawa, Yasuhiro

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Recovering a signal from its amplitude spectrogram, or phase recovery, exhibits many applications in acoustic signal processing. When only an amplitude spectrogram is available and no explicit information is given for the phases, the Griffin-Lim algorithm (GLA) is one of the most utilized methods for phase recovery. However, GLA often requires many iterations and results in low perceptual quality in some cases. In this letter, we propose two novel algorithms based on GLA and the alternating direction method of multipliers (ADMM) for better recovery with fewer iteration. Some interpretation of the existing methods and their relation to the proposed method are also provided. Evaluations are performed with both objective measure and subjective test.

AB - Recovering a signal from its amplitude spectrogram, or phase recovery, exhibits many applications in acoustic signal processing. When only an amplitude spectrogram is available and no explicit information is given for the phases, the Griffin-Lim algorithm (GLA) is one of the most utilized methods for phase recovery. However, GLA often requires many iterations and results in low perceptual quality in some cases. In this letter, we propose two novel algorithms based on GLA and the alternating direction method of multipliers (ADMM) for better recovery with fewer iteration. Some interpretation of the existing methods and their relation to the proposed method are also provided. Evaluations are performed with both objective measure and subjective test.

KW - Non-convex optimization

KW - phaseless spectrogram inversion

KW - short-time Fourier transform (STFT)

KW - spectrogram consistency

KW - STFT-based speech synthesis

UR - http://www.scopus.com/inward/record.url?scp=85057817758&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85057817758&partnerID=8YFLogxK

U2 - 10.1109/LSP.2018.2884026

DO - 10.1109/LSP.2018.2884026

M3 - Article

VL - 26

SP - 184

EP - 188

JO - IEEE Signal Processing Letters

T2 - IEEE Signal Processing Letters

JF - IEEE Signal Processing Letters

SN - 1070-9908

IS - 1

M1 - 8552369

ER -