Granted Patent B2

US 8,521,527 · App. 13/609,241 · Granted Aug 27, 2013

Abstract

Claims (67)

1. A system for processing audio in a voice response environment, comprising:

a database of host scripts each comprising signature files of audio phrases and actions to take when one of the audio phrases is recognized;

a processor to execute modules, comprising:

a generation module to generate at least one signature for each of a plurality of audio phrases, comprising:

a reference module to select a specific portion of each such audio phrase as a reference portion;

a selection module to separate the reference portion and a portion of each such audio phrase preceding the reference portion into a plurality of samples, generate a discrete Fourier transform (DFT) for each of the plurality of samples, and enable a user to select one of the DFTs for one of the samples for the reference portion as the signature for each such audio phrase based upon an auditory comparison of the samples for the reference portion with the samples preceding the reference portion; and

a file module to save the signature in the signature file of each such audio phrase;

a load module to load the host scripts and initiate a call to a voice mail server;

a buffer module to receive during the call, incoming audio buffers from voice messages stored on the voice mail server and to process the incoming audio buffers;

a signature module to create at least one signature data structure for each audio buffer;

an instruction module to maintain instruction labels, each instruction label associated with instructions to compare the signature data structure with one or more of the signatures of expected phrases in the host scripts;

a comparison module to compare the signature data structure with signatures of the expected phrases in the host scripts according to the instructions associated with at least one of the instruction labels; and

an action module to execute one or more of the actions stored in the host scripts when the signature data structure matches the signature of the expected phrase.

2. A system according to claim 1 , wherein the actions comprise at least one of sending a DTMF sequence over the call, starting audio capture and saving the captured audio into message files, playing audio files, recording one of a progress and an error message, and terminating the call.

3. A system according to claim 1 , further comprising:

a new signature module to create a new signature file for a new phrase comprising capturing audio for the new phrase, selecting the new signature for the new phrase, and editing one of the host scripts to include the new signature.

4. A system according to claim 1 , further comprising:

an association module to associate additional instructions to compare one of the signature data structures with the signatures of one or more additional expected phrases with at least one of the instruction labels.

5. A system according to claim 1 , wherein the signature comprises a two letter prefix identifying a host followed by a name of the corresponding phrase.

6. A system according to claim 1 , further comprising:

a recognition module to perform audio recognition after at least two of the audio buffers have been received.

7. A system according to claim 1 , further comprising:

a login module to log in to the voice mail server comprising identifying a security message and providing a password.

8. A system according to claim 1 , wherein the call is initiated during at least one of on demand, periodically, or based on a combination of on demand and periodically.

9. A system according to claim 1 , further comprising:

a download module to download one or more of the voice messages; and

a distribution module to distribute the one or more voice messages.

10. A system according to claim 1 , further comprising:

a display to present the voicemail messages for management and manipulation.

11. A method for processing audio in a voice response environment, comprising:

generating at least one signature for each of a plurality of audio phrases, comprising:

selecting a specific portion of each such audio phrase as a reference portion;

separating the reference portion and a portion of each such audio phrase preceding the reference portion into a plurality of samples, generating a discrete Fourier transform (DFT) for each of the plurality of samples, and enabling a user to select one of the DFTs for one of the samples for the reference portion as the signature for each such audio phrase based upon an auditory comparison of the samples for the reference portion with the samples preceding the reference portion; and

saving the signature in a signature file for each such audio phrase;

maintaining a database of host scripts each comprising the signature files of the audio phrases and actions to take when one of the audio phrases is recognized;

loading the host scripts and initiating a call to a voice mail server;

receiving during the call, incoming audio buffers from voice messages stored on the voice mail server and processing the incoming audio buffers;

creating a signature data structure for each audio buffer;

maintaining instruction labels, each instruction label associated with instructions to compare the signature data structure with one or more of the signatures of expected phrases in the host scripts;

comparing the signature data structure with the signatures of the expected phrases in the host scripts according to the instructions associated with at least one of the instruction labels; and

executing one or more of the actions stored in the host scripts when the signature data structure matches the signature of the expected phrase.

12. A method according to claim 11 , wherein the actions comprise at least one of:

sending a DTMF sequence over the call;

starting audio capture and saving the captured audio into message files;

playing audio files;

recording one of a progress and an error message; and

terminating the call.

13. A method according to claim 11 , further comprising:

creating a new signature file for a new phrase, comprising:

capturing audio for the new phrase to be learned;

selecting the new signature for the new phrase; and

editing one of the host scripts to include the new signature.

14. A method according to claim 11 , further comprising:

associating additional instructions to compare one of the signature data structures with the signatures of one or more additional expected phrases with at least one of the instruction labels.

15. A method according to claim 11 , wherein the signature comprises a two letter prefix identifying a host followed by a name of the corresponding phrase.

16. A method according to claim 11 , further comprising:

performing audio recognition after at least two of the audio buffers have been received.

17. A method according to claim 11 , further comprising:

logging in to the voice mail server comprising:

identifying a security message; and

providing a password.

18. A method according to claim 11 , wherein the call is initiated during at least one of on demand, periodically, or based on a combination of on demand and periodically.

19. A method according to claim 11 , further comprising:

downloading one or more of the voice messages; and

distributing the one or more voice messages.

20. A method according to claim 11 , further comprising:

presenting to a user, a display to access, review, manage, and manipulate the voicemail messages.

Assignments (14)