WSEAS Transactions on Signal Processing
Print ISSN: 1790-5052, E-ISSN: 2224-3488
Volume 10, 2014
Speaker Verification using Speaker-Specific-Text
Authors: ,
Abstract: In speaker recognition tasks, one of the reasons for reduced accuracy is due to closely resembling speakers in the acoustic space. In conventional GMM-based modeling technique, since the model parameters of a class are estimated without considering other classes in the system, features that are common across various classes may also be captured, along with unique features. If the system is designed to use only the unique features of a given speaker with respect to his/her acoustically resembling speaker, then the system is expected to perform better. In this proposed work, the effect of a subset of phonemes, reasonably distinct (unique) to a speaker, in the acoustic sense, on a speaker verification task is investigated. This paper proposes a technique to reduce the confusion errors, by finding speaker-specific phonemes and formulate a text using the subset of phonemes that are unique, for speaker verification task using GMM-based approach and i-vector based approach. We have experimented with three techniques namely, product of likelihood-Gaussians-based distance, Bhattacharyya distance and average loglikelihood- based distance to find out acoustically unique phonemes. Experiments have been conducted on speaker verification task using speech data of 50 speakers collected in a laboratory environment. The experiments show that the Equal Error Rate (EER) has been decreased by 4% and 4.5% using speaker-specific-text when compared to that of GMM and i-vector technique with random-text respectively.
Search Articles
Keywords: Speaker verification, Product of Gaussian, Gaussian Mixture Model, i-vector, acoustic likelihood
Pages: 320-330
WSEAS Transactions on Signal Processing, ISSN / E-ISSN: 1790-5052 / 2224-3488, Volume 10, 2014, Art. #32