Inference-aided speaker recognition

A system comprising:

• a processor configured to execute computer program instructions;
• and
• a computer storage medium encoded with computer program instructions that, when executed by the processor, cause the system to perform operations comprising:
• receiving information relating to (i) a communication device that has received an utterance and (ii) a voice associated with the received utterance, the received information relating to the communication device including a geographic location of the communication device;
• comparing the received voice information with a plurality of voice signatures in a comparison group, the comparison group corresponding to a plurality of different individuals each having an associated address within a geographic region encompassing the geographic location of the commutation device, wherein the comparison group includes at least one voice signature for each different identified individual;
• attempting to identify the voice associated with the utterance as matching the voice signature corresponding to one of the individuals in the comparison group;
• if the attempt to identify the voice associated with the utterance is unsuccessful, iteratively repeating the attempting to identify the voice associated with the utterance using a modified comparison group, the modified comparison group corresponding to a different geographic region than was used in a previous attempt to identify the voice associated with the utterance, the iterative repeating of the attempting to identify the voice associated with the utterance continuing until the voice associated with the utterance is successfully identified;
• and
• providing the communication device with access to one or more resources associated with the individual whose voice signature matched the voice associated with the utterance.

A non-transitory computer storage medium encoded with a computer program, the program comprising instructions that when executed by data processing apparatus cause the data processing apparatus to perform operations comprising:

• receiving information relating to (i) a communication device that has received an utterance and (ii) a voice associated with the received utterance, the received information relating to the communication device including a geographic location of the communication device;
• comparing the received voice information with a plurality of voice signatures in a comparison group, the comparison group corresponding to a plurality of different individuals each having an associated address within a geographic region encompassing the geographic location of the communication device, wherein the comparison group includes at least one voice signature for each different identified individual;
• attempting to identify the voice associated with the utterance as matching the voice signature corresponding to one of the individuals in the comparison group;
• if the attempt to identify the voice associated with the utterance is unsuccessful, iteratively repeating the attempting to identity the voice associated with the utterance using a modified comparison group, the modified comparison group corresponding to a different geographic region than was used in a previous attempt to identify the voice associated with the utterance, the iterative repeating of the attempting to identify the voice associated with the utterance continuing until the voice associated with the utterance is successfully identified;
• and
• providing the communication device with access to one or more resources associated with the individual whose voice signature matched the voice associated with the utterance.