A perceptual subspace approach for speech enhancement

Author

Sardaroudi, Sanaz Ghaemi ; Geravanchizadeh, Masoud

Author_Institution

Fac. of Electr. & Comput. Eng., Univ. of Tabriz, Tabriz, Iran

fYear

2010

fDate

4-6 Dec. 2010

Firstpage

878

Lastpage

881

Abstract

A Perceptual Constrained Generalized Singular Value Decomposition (PCGSVD)-based approach is proposed here for the enhancement of speech corrupted by colored noise. The proposed approach, called the PCGSVD-SDCE, uses the Spectral Domain Constrained (SDC) estimation method. Very carefully performed objective evaluations and subjective tests show that the PCGSVD-SDCE approach proposed here can offer improved speech quality and intelligibility, especially when the additive noise is nonwhite.

Keywords

singular value decomposition; speech enhancement; speech intelligibility; PCGSVD-SDCE; SDC estimation; additive noise; colored noise; perceptual constrained generalized singular value decomposition; perceptual subspace approach; spectral domain constrained estimation; speech enhancement; speech intelligibility; speech quality; Colored noise; Frequency domain analysis; Noise measurement; Speech; Speech enhancement; Auditory Masking Threshold; Perceptually constrained approach; Quotient SVD; Singular Value Decomposition; Spectral Domain Constrained Estimator; Speech Enhancement; Subspace-based Method;

fLanguage

English

Publisher

ieee

Conference_Titel

Telecommunications (IST), 2010 5th International Symposium on

Conference_Location

Tehran

Print_ISBN

978-1-4244-8183-5

Type

conf

DOI

10.1109/ISTEL.2010.5734146

Filename

5734146