Speech-Based Speaker Recognition Systems and Methods

摘要

The illustrative embodiments described herein provide systems and methods for authenticating a speaker. In one embodiment, a method includes receiving reference speech input including a reference passphrase to form a reference recording, and receiving test speech input including a test passphrase to form a test recording. The method includes determining whether the test passphrase matches the reference passphrase, and determining whether one or more voice features of the speaker of the test passphrase matches one or more voice features of the speaker of the reference passphrase. The method authenticates the speaker of the test speech input in response to determining that the reference passphrase matches the test passphrase and that one or more voice features of the speaker of the test passphrase matches one or more voice features of the speaker of the reference passphrase.

主权项

1. A method for authenticating a speaker, the method comprising:
receiving reference speech input comprising a reference passphrase to form a reference recording; determining a reference set of feature vectors for the reference recording, the reference set of feature vectors having a time dimension; receiving test speech input comprising a test passphrase to form a test recording; determining a test set of feature vectors for the test recording, the test set of feature vectors having the time dimension; correlating the reference set of feature vectors with the test set of feature vectors over the time dimension; comparing the reference set of feature vectors to the test set of feature vectors to determine whether the test passphrase matches the reference passphrase; determining a reference fundamental frequency of the reference recording; determining a test fundamental frequency of the test recording; comparing the reference fundamental frequency to the test fundamental frequency to determine whether a speaker of the test speech input matches a speaker of the reference speech input; and authenticating the speaker of the test speech input in response to determining that the reference passphrase matches the test passphrase and that the speaker of the test speech input matches the speaker of the reference speech input.