Non-intrusive codebook-based intelligibility prediction

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

2 Citationer (Scopus)

Resumé

In recent years, there has been an increasing interest in objective measures of speech intelligibility in the speech processing community. Important progress has been made in intrusive measures of intelligibility, where the Short-Time
Objective Intelligibility (STOI) method has become the de facto standard. Online adaptation of signal processing in, for example, hearing aids, in accordance with the listening conditions, requires a non-intrusive measure of intelligibility.
Presently, however, no good non-intrusive measures exist for noisy, nonstationary conditions. In this paper, we propose a novel, non-intrusive method for intelligiblity prediction in noisy conditions. The proposed method is based on STOI, which measures long-term correlations in the clean and degraded speech. Here, we propose to estimate the clean speech using a codebook-based approach that jointly models the speech and noisy spectra, parametrized by auto-regressive parameters, using pre-trained codebooks of both speech and noise. In experiments, the proposed method is demonstrated to be capable of accurately predicting the intelligibility scores obtained with STOI from oracle information. Moreover, the results are validated in listening tests that confirm that the proposed method can estimate intelligibility from noisy speech over a range of signal-to-noise ratios.
OriginalsprogEngelsk
TidsskriftSpeech Communication
Vol/bind101
Sider (fra-til)85-93
Antal sider9
ISSN0167-6393
DOI
StatusUdgivet - jul. 2018

Fingerprint

Codebook
Prediction
Speech intelligibility
Speech Intelligibility
Hearing aids
Speech Processing
Speech processing
Estimate
Signal Processing
Signal to noise ratio
Signal processing
Speech
Intelligibility
Range of data
Experiment
experiment
Experiments
community

Citer dette

@article{22515515d7d04e77bae9b966752607d2,
title = "Non-intrusive codebook-based intelligibility prediction",
abstract = "In recent years, there has been an increasing interest in objective measures of speech intelligibility in the speech processing community. Important progress has been made in intrusive measures of intelligibility, where the Short-TimeObjective Intelligibility (STOI) method has become the de facto standard. Online adaptation of signal processing in, for example, hearing aids, in accordance with the listening conditions, requires a non-intrusive measure of intelligibility.Presently, however, no good non-intrusive measures exist for noisy, nonstationary conditions. In this paper, we propose a novel, non-intrusive method for intelligiblity prediction in noisy conditions. The proposed method is based on STOI, which measures long-term correlations in the clean and degraded speech. Here, we propose to estimate the clean speech using a codebook-based approach that jointly models the speech and noisy spectra, parametrized by auto-regressive parameters, using pre-trained codebooks of both speech and noise. In experiments, the proposed method is demonstrated to be capable of accurately predicting the intelligibility scores obtained with STOI from oracle information. Moreover, the results are validated in listening tests that confirm that the proposed method can estimate intelligibility from noisy speech over a range of signal-to-noise ratios.",
keywords = "Hearing aids, non-intrusive, speech intelligibility prediction, STOI",
author = "Charlotte S{\o}rensen and Kavalekalam, {Mathew Shaji} and Angeliki Xenaki and Boldt, {Jesper B{\"u}nsow} and Christensen, {Mads Gr{\ae}sb{\o}ll}",
year = "2018",
month = "7",
doi = "10.1016/j.specom.2018.06.003",
language = "English",
volume = "101",
pages = "85--93",
journal = "Speech Communication",
issn = "0167-6393",
publisher = "Elsevier",

}

Non-intrusive codebook-based intelligibility prediction. / Sørensen, Charlotte; Kavalekalam, Mathew Shaji; Xenaki, Angeliki; Boldt, Jesper Bünsow; Christensen, Mads Græsbøll.

I: Speech Communication, Bind 101, 07.2018, s. 85-93.

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

TY - JOUR

T1 - Non-intrusive codebook-based intelligibility prediction

AU - Sørensen, Charlotte

AU - Kavalekalam, Mathew Shaji

AU - Xenaki, Angeliki

AU - Boldt, Jesper Bünsow

AU - Christensen, Mads Græsbøll

PY - 2018/7

Y1 - 2018/7

N2 - In recent years, there has been an increasing interest in objective measures of speech intelligibility in the speech processing community. Important progress has been made in intrusive measures of intelligibility, where the Short-TimeObjective Intelligibility (STOI) method has become the de facto standard. Online adaptation of signal processing in, for example, hearing aids, in accordance with the listening conditions, requires a non-intrusive measure of intelligibility.Presently, however, no good non-intrusive measures exist for noisy, nonstationary conditions. In this paper, we propose a novel, non-intrusive method for intelligiblity prediction in noisy conditions. The proposed method is based on STOI, which measures long-term correlations in the clean and degraded speech. Here, we propose to estimate the clean speech using a codebook-based approach that jointly models the speech and noisy spectra, parametrized by auto-regressive parameters, using pre-trained codebooks of both speech and noise. In experiments, the proposed method is demonstrated to be capable of accurately predicting the intelligibility scores obtained with STOI from oracle information. Moreover, the results are validated in listening tests that confirm that the proposed method can estimate intelligibility from noisy speech over a range of signal-to-noise ratios.

AB - In recent years, there has been an increasing interest in objective measures of speech intelligibility in the speech processing community. Important progress has been made in intrusive measures of intelligibility, where the Short-TimeObjective Intelligibility (STOI) method has become the de facto standard. Online adaptation of signal processing in, for example, hearing aids, in accordance with the listening conditions, requires a non-intrusive measure of intelligibility.Presently, however, no good non-intrusive measures exist for noisy, nonstationary conditions. In this paper, we propose a novel, non-intrusive method for intelligiblity prediction in noisy conditions. The proposed method is based on STOI, which measures long-term correlations in the clean and degraded speech. Here, we propose to estimate the clean speech using a codebook-based approach that jointly models the speech and noisy spectra, parametrized by auto-regressive parameters, using pre-trained codebooks of both speech and noise. In experiments, the proposed method is demonstrated to be capable of accurately predicting the intelligibility scores obtained with STOI from oracle information. Moreover, the results are validated in listening tests that confirm that the proposed method can estimate intelligibility from noisy speech over a range of signal-to-noise ratios.

KW - Hearing aids, non-intrusive, speech intelligibility prediction, STOI

UR - http://www.scopus.com/inward/record.url?scp=85049329391&partnerID=8YFLogxK

U2 - 10.1016/j.specom.2018.06.003

DO - 10.1016/j.specom.2018.06.003

M3 - Journal article

VL - 101

SP - 85

EP - 93

JO - Speech Communication

JF - Speech Communication

SN - 0167-6393

ER -