Generalized Discriminant Methods for Improved X-Vector Back-end Based Stress Speech Recognition

カテゴリ: 論文誌(論文単位)

グループ名: 【C】電子・情報・システム部門

発行日: 2019/11/01

タイトル(英語): Generalized Discriminant Methods for Improved X-Vector Back-end Based Stress Speech Recognition

著者名: Barlian Henryranu Prasetio (Interdisciplinary Graduate School of Agriculture and Engineering, University of Miyazaki), Hiroki Tamura (Faculty of Engineering, University of Miyazaki), Koichi Tanno (Faculty of Engineering, University of Miyazaki)

著者名(英語): Barlian Henryranu Prasetio (Interdisciplinary Graduate School of Agriculture and Engineering, University of Miyazaki), Hiroki Tamura (Faculty of Engineering, University of Miyazaki), Koichi Tanno (Faculty of Engineering, University of Miyazaki)

キーワード: stress speech，X-Vector，LDA，GDA，PLDA，joint-PLDA

要約(英語): In this paper, we discuss a system that can recognize and classify stress based on speech. The system was built using a deep neural network (DNN) embedding to improve the baseline system (i-vector), called x-vector. The i-vector system usually requires Linear Discriminant Analysis (LDA) for dimension reduction, followed by Probability Linear Discriminant Analysis (PLDA) for scoring. In this work, the Generalized Discriminant Analysis (GDA) based Gaussianized cosine kernel and Joint-PLDA was used to replace the LDA and PLDA, respectively. The SUSAS database was used for training, testing, and enrollment data of our proposed system. We assessed our proposed system's effectiveness compared to the baseline system using equal error rate (EER). The evaluation and experiment results show that our proposed system outperforms the baseline system.

本誌: 電気学会論文誌C（電子・情報・システム部門誌） Vol.139 No.11 （2019）特集Ⅰ：国際会議ICESS 2018　特集Ⅱ：電気関係学会関西連合大会

本誌掲載ページ: 1341-1347 p

原稿種別: 論文／英語

電子版へのリンク: https://www.jstage.jst.go.jp/article/ieejeiss/139/11/139_1341/_article/-char/ja/