WEKO3
-
RootNode
アイテム
Inter-speaker weighted MAP adaptation for GMM-supervector speaker recognition
https://ipsj.ixsq.nii.ac.jp/records/71574
https://ipsj.ixsq.nii.ac.jp/records/71574bbb96bda-c31f-44c9-8679-91b044cb7fac
名前 / ファイル | ライセンス | アクション |
---|---|---|
![]() |
Copyright (c) 2010 by the Information Processing Society of Japan
|
|
オープンアクセス |
Item type | SIG Technical Reports(1) | |||||||
---|---|---|---|---|---|---|---|---|
公開日 | 2010-12-13 | |||||||
タイトル | ||||||||
タイトル | Inter-speaker weighted MAP adaptation for GMM-supervector speaker recognition | |||||||
タイトル | ||||||||
言語 | en | |||||||
タイトル | Inter-speaker weighted MAP adaptation for GMM-supervector speaker recognition | |||||||
言語 | ||||||||
言語 | eng | |||||||
キーワード | ||||||||
主題Scheme | Other | |||||||
主題 | Session-4 話者認識・識別 | |||||||
資源タイプ | ||||||||
資源タイプ識別子 | http://purl.org/coar/resource_type/c_18gh | |||||||
資源タイプ | technical report | |||||||
著者所属 | ||||||||
Tokyo Institute of Technology | ||||||||
著者所属 | ||||||||
Tokyo Institute of Technology | ||||||||
著者所属 | ||||||||
Tokyo Institute of Technology | ||||||||
著者所属(英) | ||||||||
en | ||||||||
Tokyo Institute of Technology | ||||||||
著者所属(英) | ||||||||
en | ||||||||
Tokyo Institute of Technology | ||||||||
著者所属(英) | ||||||||
en | ||||||||
Tokyo Institute of Technology | ||||||||
著者名 |
Marc, Ferras
Koichi, Shinoda
Sadaoki, Furui
× Marc, Ferras Koichi, Shinoda Sadaoki, Furui
|
|||||||
著者名(英) |
Marc, Ferras
Koichi, Shinoda
Sadaoki, Furui
× Marc, Ferras Koichi, Shinoda Sadaoki, Furui
|
|||||||
論文抄録 | ||||||||
内容記述タイプ | Other | |||||||
内容記述 | Gaussian Mixture Models (GMM) are ubiquitously used in state-of-the-art speaker recognition systems. The popular GMM-SVM paradigm uses Maximum A Posteriori (MAP) speaker-adapted GMM models by stacking the mean vectors into a supervector that is fed into a Support Vector Machine classifier. In this paper, we modify the standard relevance MAP algorithm to better fit the speaker recognition task. We propose to emphasize the adaptation of the Gaussian mixtures according to the inter-speaker variability exhibited on a training set, thus accounting for both the occupation count and the speaker discrimination ability during adaptation. We evaluate our proposal on a relevance MAP based GMM-SVM system using a large telephone speech corpus such as the one provided in the 2006 NIST Speaker Recognition Evaluation. We show that despite its simplicity this technique is effective. | |||||||
論文抄録(英) | ||||||||
内容記述タイプ | Other | |||||||
内容記述 | Gaussian Mixture Models (GMM) are ubiquitously used in state-of-the-art speaker recognition systems. The popular GMM-SVM paradigm uses Maximum A Posteriori (MAP) speaker-adapted GMM models by stacking the mean vectors into a supervector that is fed into a Support Vector Machine classifier. In this paper, we modify the standard relevance MAP algorithm to better fit the speaker recognition task. We propose to emphasize the adaptation of the Gaussian mixtures according to the inter-speaker variability exhibited on a training set, thus accounting for both the occupation count and the speaker discrimination ability during adaptation. We evaluate our proposal on a relevance MAP based GMM-SVM system using a large telephone speech corpus such as the one provided in the 2006 NIST Speaker Recognition Evaluation. We show that despite its simplicity this technique is effective. | |||||||
書誌レコードID | ||||||||
収録物識別子タイプ | NCID | |||||||
収録物識別子 | AN10442647 | |||||||
書誌情報 |
研究報告音声言語情報処理(SLP) 巻 2010-SLP-84, 号 12, p. 1-4, 発行日 2010-12-13 |
|||||||
Notice | ||||||||
SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc. | ||||||||
出版者 | ||||||||
言語 | ja | |||||||
出版者 | 情報処理学会 |