A novel alternative hypothesis characterization using kernel classifiers for LLR-based speaker verification
檔案
本條目沒有附加檔案。
日期
2006
期刊名稱
CHINESE SPOKEN LANGUAGE PROCESSING, PROCEEDINGS
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.