Discriminatory and Orthogonal Feature Learning for Noise Robust Keyword Spotting
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Kim, Donghyeon | - |
dc.contributor.author | Ko, Kyungdeuk | - |
dc.contributor.author | Han, David K. | - |
dc.contributor.author | Ko, Hanseok | - |
dc.date.accessioned | 2022-11-16T03:40:42Z | - |
dc.date.available | 2022-11-16T03:40:42Z | - |
dc.date.created | 2022-11-15 | - |
dc.date.issued | 2022 | - |
dc.identifier.issn | 1070-9908 | - |
dc.identifier.uri | https://scholar.korea.ac.kr/handle/2021.sw.korea/145567 | - |
dc.description.abstract | Keyword Spotting (KWS) is an essential component in a smart device for alerting the system when a user prompts it with a command. As these devices are typically constrained by computational and energy resources, the KWS model should be designed with a small footprint. In our previous work, we developed lightweight dynamic filters which extract a robust feature map within a noisy environment. The learning variables of the dynamic filter are jointly optimized with KWS weights by using Cross-Entropy (CE) loss. CE loss alone, however, is not sufficient for high performance when the SNR is low. In order to train the network for more robust performance in noisy environments, we introduce the LOw Variant Orthogonal (LOVO) loss. The LOVO loss is composed of a triplet loss applied on the output of the dynamic filter, a spectral norm-based orthogonal loss, and an inner class distance loss applied in the KWS model. These losses are particularly useful in encouraging the network to extract discriminatory features in unseen noise environments. | - |
dc.language | English | - |
dc.language.iso | en | - |
dc.publisher | IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC | - |
dc.title | Discriminatory and Orthogonal Feature Learning for Noise Robust Keyword Spotting | - |
dc.type | Article | - |
dc.contributor.affiliatedAuthor | Ko, Hanseok | - |
dc.identifier.doi | 10.1109/LSP.2022.3203911 | - |
dc.identifier.scopusid | 2-s2.0-85137881036 | - |
dc.identifier.wosid | 000853834100005 | - |
dc.identifier.bibliographicCitation | IEEE SIGNAL PROCESSING LETTERS, v.29, pp.1913 - 1917 | - |
dc.relation.isPartOf | IEEE SIGNAL PROCESSING LETTERS | - |
dc.citation.title | IEEE SIGNAL PROCESSING LETTERS | - |
dc.citation.volume | 29 | - |
dc.citation.startPage | 1913 | - |
dc.citation.endPage | 1917 | - |
dc.type.rims | ART | - |
dc.type.docType | Article | - |
dc.description.journalClass | 1 | - |
dc.description.journalRegisteredClass | scie | - |
dc.description.journalRegisteredClass | scopus | - |
dc.relation.journalResearchArea | Engineering | - |
dc.relation.journalWebOfScienceCategory | Engineering, Electrical & Electronic | - |
dc.subject.keywordAuthor | Measurement | - |
dc.subject.keywordAuthor | Computational modeling | - |
dc.subject.keywordAuthor | Feature extraction | - |
dc.subject.keywordAuthor | Mathematical models | - |
dc.subject.keywordAuthor | Convolution | - |
dc.subject.keywordAuthor | Training | - |
dc.subject.keywordAuthor | Euclidean distance | - |
dc.subject.keywordAuthor | Keyword Spotting | - |
dc.subject.keywordAuthor | robustness | - |
dc.subject.keywordAuthor | metric learning | - |
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.
145 Anam-ro, Seongbuk-gu, Seoul, 02841, Korea+82-2-3290-2963
COPYRIGHT © 2021 Korea University. All Rights Reserved.
Certain data included herein are derived from the © Web of Science of Clarivate Analytics. All rights reserved.
You may not copy or re-distribute this material in whole or in part without the prior written consent of Clarivate Analytics.