A novel alternative hypothesis characterization using kernel classifiers for LLR-based speaker verification
顯示簡易詮釋資料 統計資料| 欄位 | 內容 | 語言 |
|---|---|---|
dc.citation.epage |
+ | en_US |
dc.citation.spage |
506 | en_US |
dc.citation.volume |
4274 | en_US |
dc.citation.woscount |
0 | en_US |
dc.contributor.author |
Chao, Yi-Hsiang | en_US |
dc.contributor.author |
Wang, Hsin-Min | en_US |
dc.contributor.author |
Chang, Ruei-Chuan | en_US |
dc.contributor.department |
資訊工程學系 | zh_TW |
dc.contributor.department |
Department of Computer Science | en_US |
dc.date.accessioned |
2017-04-21T06:49:40Z | |
dc.date.available |
2017-04-21T06:49:40Z | |
dc.date.issued |
2006 | en_US |
dc.description.abstract |
In a log-likelihood ratio (LLR)-based speaker verification system, the alternative hypothesis is usually ill-defined and hard to characterize a priori, since it should cover the space of all possible impostors. In this paper, we propose a new LLR measure in an attempt to characterize the alternative hypothesis in a more effective and robust way than conventional methods. This LLR measure can be further formulated as a non-linear discriminant classifier and solved by kernel-based techniques, such as the Kernel Fisher Discriminant (KFD) and Support Vector Machine (SVM). The results of experiments on two speaker verification tasks show that the proposed methods outperform classical LLR-based approaches. | en_US |
dc.identifier.isbn |
978-3-540-49665-6 | en_US |
dc.identifier.issn |
0302-9743 | en_US |
dc.identifier.journal |
CHINESE SPOKEN LANGUAGE PROCESSING, PROCEEDINGS | en_US |
dc.identifier.uri |
https://ir.lib.nycu.edu.tw/handle/11536/134489 | |
dc.identifier.wosnumber |
WOS:000244824800048 | en_US |
dc.language.iso |
en_US | en_US |
dc.subject |
speaker verification | en_US |
dc.subject |
log-likelihood ratio | en_US |
dc.subject |
Kernel Fisher Discriminant | en_US |
dc.subject |
Support Vector Machine | en_US |
dc.title |
A novel alternative hypothesis characterization using kernel classifiers for LLR-based speaker verification | en_US |
dc.type |
Proceedings Paper | en_US |
檔案
本條目沒有附加檔案。
所屬典藏: 會議論文
引用此條目
Chao, Y.-H., Wang, H.-M., & Chang, R.-C. (2006). A novel alternative hypothesis characterization using kernel classifiers for LLR-based speaker verification. CHINESE SPOKEN LANGUAGE PROCESSING, PROCEEDINGS, 4274, 506–+. https://ir.lib.nycu.edu.tw/handle/11536/134489
引用文字依 APA 第 7 版格式自動產生,使用前請再核對。
分享
永久網址:https://ir.lib.nycu.edu.tw/handle/11536/134489