Closed-loop command and response system for automatic communications between interacting computer systems over an audio communications channel
View Patent ↗A system and method for enabling two computer systems to communicate over an audio communications channel, such as a voice telephony connection. Such a system includes a software application that enables a user's computer to call, interrogate, download, and manage a voicemail account stored on a telephone company's computer, without human intervention. A voicemail retrieved from the telephone company's computer can be stored in a digital format on the user's computer. In such a format, the voicemail can be readily archived, or even distributed throughout a network, such as the Internet, in a digital form, such as an email attachment. Preferably a computationally efficient audio recognition algorithm is employed by the user's computer to respond to and navigate the automated audio menu of the telephone company's computer.
1. A computationally efficient method for recognizing audio signals, comprising the steps of:
generating a plurality of known discrete Fourier transforms (DFTs), each known DFT generated from a different specific audio signal, comprising:
determining the audio signals that are likely to be received; and
generating a known DFT for each audio signal, comprising:
selecting a specific portion of each such audio signal as a reference portion; and
creating a signature for each such audio signal by separating the reference portion and a portion of the audio signal preceding the reference portion into a plurality of samples, generating a DFT for each of the plurality of samples, and enabling a user to select one DFT from the plurality of samples in the reference portion as the known DFT based upon an auditory comparison of the samples for the reference portion with the samples preceding the reference portion;
receiving an audio signal;
generating at least one unknown DFT from the audio signal received; and
comparing the at least one unknown DFT generated from the audio signal received, with the known DFTs, to identify the audio signal that was received based on a best match of the unknown DFT to one of the known DFTs.
2. The method of claim 1 , further comprising the steps of:
storing the audio signal received in an audio buffer; and
separating contents of the audio buffer into a plurality of equally sized samples, before generating the at least one unknown DFT, wherein the step of generating the at least one unknown DFT comprises the step of generating an unknown DFT for each sample, thereby generating a plurality of unknown DFTs that are then compared to the known DFTs to determine the best match.
3. The method of claim 1 , further comprising the step of executing a predefined action associated with said one of the known DFTs that was identified in the comparison of the at least one unknown DFT with the known DFTs.
4. The method of claim 3 , wherein a different predefined action is associated with each known DFT.
5. The method of claim 3 , wherein the predefined action comprises at least one step selected from the group of steps consisting of:
initiating a telephone call;
inputting a password;
playing a message;
returning a number of available messages to access;
terminating a telephone call;
recording a message;
saving a message; and
recording a message in a log file.
6. The method of claim 1 , wherein the step of generating a known DFT for each such audio signal comprises capturing each such audio signal.
7. The method of claim 1 , wherein selecting a specific portion of each such audio signal further comprises enabling the user to listen to the selected reference portion of each such audio signal to determine if the selected reference portion is acceptable.
8. The method of claim 1 , further comprising:
enabling the user to select a different specific portion to serve as a different reference portion if the user is not satisfied with the comparison of the samples from the reference portion and the samples preceding the reference portion; and
separating the different reference portion into a plurality of different samples;
generating the DFT for each of the different samples in the different reference portion; and
enabling the user to select one DFT from the different samples in the different reference portion, until the user is satisfied with the comparison.
9. The method of claim 1 , further comprising:
analyzing an audio spectrum of the signature to determine if a clean segment of speech, which is relatively free of extraneous noise, has been selected for the reference portion.
10. The method of claim 9 , further comprising:
inspecting at least one correlation variable that is superimposed proximate to the audio spectrum.
11. A system for recognizing an audio signal, comprising:
a memory, comprising:
a plurality of machine instructions defining an audio recognition program; and
a plurality of known discrete Fourier transforms (DFTs), each known DFT corresponding to a different specific audio signal; and
a processor that is coupled to the memory, to access the machine instructions and the known DFTs, said processor executing the machine instructions and thereby implementing a plurality of functions, including:
receiving an audio signal;
generating at least one unknown DFT based on the audio signal received; and
comparing the at least one unknown DFT with the known DFTs, to identify the audio signal received based on selecting one of the known DFTs with the best match and to determine the specific audio signal corresponding to the known DFTs with the best match; and
a display coupled to the processor, wherein the display includes a graphical user interface that includes a plurality of controls configured to enable a user to generate the plurality of known DFTs, each generated from one of the specific audio signals.
12. The system of claim 11 , further comprising:
a voice modem coupled to the processor and the memory to enable capture of each different specific audio signal, by passing digitized audio signals to the processor.
13. The system of claim 11 , further comprising:
a speaker coupled to the processor to enable a user to listen to an audio segment of the different specific audio signal.
14. The system of claim 11 , wherein the graphical user interface is configured to display the different specific audio signal as an audio amplitude waveform.
15. The system of claim 11 , wherein at least one control is configured to enable the user to select a specific portion of the different specific audio signal as a reference portion.
16. The system of claim 15 , wherein at least one control is configured to enable the user to create the known DFT of the reference portion, and further wherein execution of the machine instructions comprise the method of:
separating the reference portion and a portion of the different specific audio signals preceding the reference portion into a plurality of samples;
generating a DFT for each of the plurality of samples; and
selecting one DFT for one of the samples corresponding to the reference portion as the known DFT, based upon a comparison of the samples for the reference portion with the samples preceding the reference portion.
17. The system of claim 13 , wherein at least one control is configured to enable the user to selectively manipulate the playing of the audio segment.
18. The system of claim 11 , further comprising a plurality of buffers, each buffer configured to store portions of at least one sample of a signal selected from the group consisting of:
the audio signal;
the unknown DFT; and
the known DFT.
19. A non-transitory computer-readable storage on which machine-readable instructions are stored, which when executed by a computing device cause the computing device to carry out a method, comprising:
generating a plurality of known discrete Fourier transforms (DFTs), each known DFT generated from a different specific audio signal, comprising:
determining the audio signals that are likely to be received; and
generating a known DFT for each audio signal, comprising:
selecting a specific portion of each such audio signal as a reference portion; and
creating a signature for each such audio signal by separating the reference portion and the portion of the audio preceding the reference portion into a plurality of samples, generating a DFT for each of the plurality of samples, and enabling a user to select one DFT from the plurality of samples in the reference portion as the known DFT based upon an auditory comparison of the samples for the reference portion with the samples preceding the reference portion;
receiving an audio signal;
generating at least one unknown DFT from the audio signal received; and
comparing the at least one unknown DFT generated from the audio signal received, with the known DFTs, to identify the audio signal that was received based on a best match of the unknown DFT to one of the known DFTs.