Frequency Offset Correction in Single Sideband (SSB) Speech by Deep Neural Network for Speaker Verification

DSpace/Manakin Repository

Frequency Offset Correction in Single Sideband (SSB) Speech by Deep Neural Network for Speaker Verification

Show full item record

Title: Frequency Offset Correction in Single Sideband (SSB) Speech by Deep Neural Network for Speaker Verification
Author(s):
Xing, Hua (UT Dallas);
Liu, Gang (UT Dallas);
Hansen, John H. L. (UT Dallas)
Item Type: article
Keywords: Show Keywords
Abstract: Communication system mismatch represents a major influence for loss in speaker recognition performance. This paper considers a type of nonlinear communication system mismatch- modulation/ demodulation (Mod/DeMod) carrier drift in single sideband (SSB) speech signals. We focus on the problem of estimating frequency offset in SSB speech in order to improve speaker verification performance of the drifted speech. Based on a two-step framework from previous work, we propose using a multi-layered neural network architecture, stacked denoising autoencoder (SDA), to determine the unique interval of the offset value in the first step. Experimental results demonstrate that the SDA based system can produce up to a +16.1% relative improvement in frequency offset estimation accuracy. A speaker verification evaluation shows a +65.9% relative improvement in EER when SSB speech signal is compensated with the frequency offset value estimated by the proposed method.
Publisher: International Speech and Communication Association
ISSN: 2308-457X (ISSN)
Persistent Link: http://www.isca-speech.org/archive/interspeech_2015/i15_1156.html
http://hdl.handle.net/10735.1/5108
Bibliographic Citation: Xing, H., G. Liu, and J. H. L. Hansen. 2015. "Frequency offset correction in single sideband (SSB) speech by deep neural network for speaker verification." INTERSPEECH 2015 (16th Annual Conference of the International Speech Communication Association) p. 1156-1160.
Terms of Use: ©2015
Sponsors:

Files in this item

Files Size Format View
JECS-3626-4714.34.pdf 128.1Kb PDF View/Open Article

This item appears in the following Collection(s)


Show full item record