Non-intrusive Speech Quality Assessment based on Tucker Decomposition and Deep Neural Network

Yahui Shan, Jing Wang, Min Liu, Yiyu Luo, Xiang Xie

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

In this paper, we present a new speech quality assessment method to estimate the quality of degraded speech without the reference speech. The traditional non-intrusive assessment methods cannot meet the requirement of high consistency with subjective results owing to the lack of original reference signals. To solve these issues, deep belief network is trained to produce pseudo-reference speech signal of degraded speech. Then the pseudo-reference speech and the degraded speech are modeled by tensor analysis to obtained features which is used to calculate feature differences. The feature differences are mapped to speech quality score using support vector regression. Experiments are conducted in a wideband dataset containing various degraded speech signals and subjective listening scores. When compared with the Gaussian Mixture Model method and deep belief network method, the proposed method brings about a higher correlation coefficient between predicted scores and subjective scores.

Original languageEnglish
Title of host publicationICSIDP 2019 - IEEE International Conference on Signal, Information and Data Processing 2019
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781728123455
DOIs
Publication statusPublished - Dec 2019
Event2019 IEEE International Conference on Signal, Information and Data Processing, ICSIDP 2019 - Chongqing, China
Duration: 11 Dec 201913 Dec 2019

Publication series

NameICSIDP 2019 - IEEE International Conference on Signal, Information and Data Processing 2019

Conference

Conference2019 IEEE International Conference on Signal, Information and Data Processing, ICSIDP 2019
Country/TerritoryChina
CityChongqing
Period11/12/1913/12/19

Keywords

  • deep belief network
  • non-intrusive speech quality assessment
  • support vector regression
  • tensor analysis

Fingerprint

Dive into the research topics of 'Non-intrusive Speech Quality Assessment based on Tucker Decomposition and Deep Neural Network'. Together they form a unique fingerprint.

Cite this