Title :
Speaker Identification by Multi-Frame Generative Models
Author :
Impedovo, Donato ; Refice, Mario
Author_Institution :
Dipt. di Elettrotec. ed Elettron., Politec. di Bari, Bari
Abstract :
In this paper an approach called multi-frame speaker models (MFS) is proposed, in order to cope with performance degradation generally observed over (short and medium) time and trials in speaker identification´s task. The approach, based on generative models, uses multiple frame´s length for speech processing in training and testing phase. A complete multi-expert system is also presented which is able to implement the proposed approach onthe whole set of speakers and to obtain a near optimum for the ER´s reduction.
Keywords :
expert systems; speaker recognition; ER reduction; multiexpert system; multiframe generative model; multiframe speaker model; speaker identification; speech processing; Biometrics; Character recognition; Degradation; Erbium; Feature extraction; Hidden Markov models; Impedance; Information security; Speech; Testing; Biometrics; Multi Expert; Multi-Frame; Speaker Identification;
Conference_Titel :
Information Assurance and Security, 2008. ISIAS '08. Fourth International Conference on
Conference_Location :
Naples
Print_ISBN :
978-0-7695-3324-7
DOI :
10.1109/IAS.2008.15