Method and System for Pith Synchronous Feature Generation of Speaker Recognition System

摘要

The present invention relates to a speaker recognition system related to a speaker's feature vector generation method and system of the speaker recognition system extracts a feature vector from the input talker speech signal to perform the speaker model training and speaker recognition, the input talker speech signal to measure the pitch and pitch period made, including voice extracting means, the feature vector generating means for generating a characteristic vector of a speech signal extracted from the voice pitch period extraction means for extracting a predetermined pitch period speech signal is characterized.