A Proposed Decision Rule for Speaker Identification Based on a Posteriori Probability
نویسندگان
چکیده
In speaker recognition, the maximum likelihood (ML) rule is used as a criterion to assign a given sequence of acoustic vectors to the maximum likelihood speaker model. However, this rule is not flexible in some cases. An alternative decision rule, the maximum average normalised likelihood (MANL), is proposed in this paper. The theoretical analysis and the experimental results show that the MANL rule can be used in speaker identification and it is more effective than the ML rule in the approaches based on Gaussian mixture model (GMM) and vector quantisation (VQ).
منابع مشابه
Segmental Normalization for Robust Speaker Verification
For the task of speaker verification, similarity measure normalization methods are relevant to cope with variability problems and with data and/or decision fusion problems. The aim of this paper is to suggest a new method of normalization which combines classical world model based normalization techniques with ones based on a posteriori probability. This original method presents the well-known ...
متن کاملA proposed decision rule for speaker recognition based on fuzzy c-means clustering
In vector quantisation (VQ) based speaker recognition, the minimum overall average distortion rule is used as a criterion to assign a given sequence of acoustic vectors to a speaker model known as a codebook. An alternative decision rule based on fuzzy c-means clustering is proposed in this paper. A set of membership functions associated with vectors for codebooks are defined as discriminant fu...
متن کاملSpeaker Identification by Joint Statistical Characterization in the Log Gabor Wavelet Domain
Real world Speaker Identification (SI) application differs from ideal or laboratory conditions causing perturbations that leads to a mismatch between the training and testing environment and degrade the performance drastically. Many strategies have been adopted to cope with acoustical degradation; wavelet based Bayesian marginal model is one of them. But Bayesian marginal models cannot model th...
متن کاملSimilarity normalization method based on world model and a posteriori probability for speaker verification
For the task of speaker verification, similarity measure normalization methods are relevant to cope with variability problems and with data and/or decision fusion issues. The aim of this paper is to suggest a new normalization method, which combines classical world model-based normalization techniques with a posteriori probability-based ones. This method presents the wellknown advantages of the...
متن کاملMMDT: Multi-Objective Memetic Rule Learning from Decision Tree
In this article, a Multi-Objective Memetic Algorithm (MA) for rule learning is proposed. Prediction accuracy and interpretation are two measures that conflict with each other. In this approach, we consider accuracy and interpretation of rules sets. Additionally, individual classifiers face other problems such as huge sizes, high dimensionality and imbalance classes’ distribution data sets. This...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 1998