AUTOMATIC TEXT-INDEPENDENT, LANGUAGE-INDEPENDENT SPEAKER VOICE-PRINT CREATION AND SPEAKER RECOGNITION

摘要

Disclosed herein is an automatic dual-step, text~ independent, language-independent speaker voice-print creation and speaker recognition method,wherein a neural network-based technique is used in a first step and a Markovmodel-based technique is used in the second step. In particular, the firststep uses a neural network-based technique for decoding the content of what isuttered by the speaker in terms of language~ independent acoustic-phoneticclasses, wherein the second step uses the sequence of language-independentacoustic-phonetic classes from the first step and employs a Markov model-basedtechnique for creating the speaker voice-print and for recognizing thespeaker. The combination of the two steps enables improvement in the accuracyand efficiency of the speaker voice-print creation and of the speakerrecognition, without setting any constraints on the lexical content of thespeaker utterance and on the language thereof.