標題: A novel alternative hypothesis characterization using kernel classifiers for LLR-based speaker verification
作者: Chao, Yi-Hsiang
Wang, Hsin-Min
Chang, Ruei-Chuan
資訊工程學系
Department of Computer Science
關鍵字: speaker verification;log-likelihood ratio;Kernel Fisher Discriminant;Support Vector Machine
公開日期: 2006
摘要: 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.
URI: http://hdl.handle.net/11536/134489
ISBN: 978-3-540-49665-6
ISSN: 0302-9743
期刊: CHINESE SPOKEN LANGUAGE PROCESSING, PROCEEDINGS
Volume: 4274
起始頁: 506
結束頁: +
Appears in Collections:Conferences Paper