Speaker verification methods and apparatus

At least one computer-readable storage device storing at least one program for execution on at least one processor, the program having instructions that, when executed on the at least one processor, perform a method for determining a validity of an identity asserted by a speaker using a voice print that models speech of a user whose identity the speaker is asserting, the method comprising acts of:

• performing a first verification stage comprising acts of:
• obtaining a first voice signal from the speaker uttering at least one first challenge utterance;
• and
• comparing at least one characteristic feature of the first voice signal with at least a portion of the voice print to assess whether the at least one characteristic feature of the first voice signal is similar enough to the at least a portion of the voice print to conclude that the first voice signal was obtained from an utterance by the user;
• and
• performing a second verification stage in a same session as the first verification stage if it is concluded in the first verification stage that the first voice signal was obtained from an utterance by the user, the second verification stage comprising acts of:
• adapting at least one parameter of the voice print based, at least in part, on the first voice signal to obtain an adapted voice print;
• obtaining a second voice signal from the speaker uttering at least one second challenge utterance;
• and
• comparing at least one characteristic feature of the second voice signal with at least a portion of the adapted voice print to assess whether the at least one characteristic feature of the second voice signal is similar enough to the at least a portion of the adapted voice print to conclude that the second voice signal was obtained from an utterance by the user.

A speaker verification system comprising:

• at least one computer readable storage medium storing at least one voice print, the at least one voice print obtained from characteristic features of at least one first voice signal obtained from a user uttering at least one enrollment utterance;
• a receiver to receive voice signals of speakers being verified responding to prompts to utter one or more challenge utterances;
• and
• at least one controller coupled to the at least one computer readable storage medium and the receiver, the at least one controller configured to perform a first verification stage comprising:
• obtaining from the receiver a first voice signal from a speaker who is being verified, the first voice signal obtained from the speaker uttering at least one first challenge utterance;
• and
• comparing at least one characteristic feature of the first voice signal with at least a portion of a first voice print of the at least one voice print obtained from the user whose identity the speaker is asserting to assess whether the at least one characteristic feature of the first voice signal is similar enough to the at least a portion of the first voice print to conclude that the first voice signal was obtained from an utterance by the user,
• wherein the at least one controller is configured to perform a second verification stage in a same session as the first verification stage if it is concluded in the first verification stage that the first voice signal was obtained from an utterance by the user, the at least one controller, the second verification stage comprising:
• adapting at least one parameter of the first voice print based, at least in part, on the first voice signal to obtain an adapted voice print;
• obtaining from the receiver a second voice signal from the speaker uttering at least one second challenge utterance;
• and
• comparing at least one characteristic feature of the second voice signal with at least a portion of the adapted voice print to assess whether the at least one characteristic feature of the second voice signal is similar enough to the at least a portion of the adapted voice print to conclude that the second voice signal was obtained from an utterance by the user.