Single-Channel Target Speaker Extraction System with Attention Enhancement

Yen Ting Lai, Yi En Lin, Pao Chi Chang, Jia Ching Wang

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

Abstract

In this paper, we propose a system for single-channel target speaker extraction. We adopt a Temporal Convolutional Network (TCN) architecture as speech extraction model. We also import an attention enhancement to provide system more rich and efficient information. This can improve the extraction model to better estimate the mask of the target. With the better mask, the quality of the target speaker extraction is noticeably improved.

Original languageEnglish
Title of host publicationProceedings - 2022 IEEE International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2022
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages433-434
Number of pages2
ISBN (Electronic)9781665470506
DOIs
StatePublished - 2022
Event2022 IEEE International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2022 - Taipei, Taiwan
Duration: 6 Jul 20228 Jul 2022

Publication series

NameProceedings - 2022 IEEE International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2022

Conference

Conference2022 IEEE International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2022
Country/TerritoryTaiwan
CityTaipei
Period6/07/228/07/22

Fingerprint

Dive into the research topics of 'Single-Channel Target Speaker Extraction System with Attention Enhancement'. Together they form a unique fingerprint.

Cite this