Method and apparatus for using the control channel in telecommunications systems for voice dialing

• US 6,236,715 B1
• Filed: 04/15/1997
• Issued: 05/22/2001
• Est. Priority Date: 04/15/1997
• Status: Expired due to Term

• Current Assignee: RPX Clearinghouse LLC (RPX Corporation)
• Original Assignee: Nortel Networks Corporation
• Inventors: Wong, Chi
• Primary Examiner(s): WEAVER, SCOTT LOUIS
• Application Number:
  US08/842,605
• Time in Patent Office: 1,498 Days
• Field of Search: 379/88.01, 379/88.02, 379/88.03, 379/88.04, 379/88.07, 379/88.1, 379/88.17, 704/270, 704/275, 370/524, 455/563
• US Class Current: 379/88.03
• CPC Class Codes: H04M 2207/08 ISDN systems
• CPC Class Codes: H04M 3/42204 Arrangements at the exchange for service or number selection by voice at the terminal H04M1/27
• CPC Class Codes: H04M 3/42314 in private branch exchanges

Abstract

A voice-dialing apparatus and method extracts semantic features from a speech signal and transmits a set of semantic feature characteristics over a telecommunication control channel to a remote speech recognition processor. Once received, the remote speech recognition processor uses the feature set to retrieve a name and number corresponding to the audio signal received from the user and to automatically dial the number for the user.

Claims

1. A voice-activated dialing method, comprising the steps of:

receiving, at a telephone terminal, an audio signal corresponding to a desired destination;

locally conditioning the audio signal to extract a set of semantic feature characteristics;

locally forming a feature signal representing the set of semantic feature characteristics corresponding to the received audio signal;

transmitting the feature signal via a control channel to a remote speech recognition processor;

retrieving, with the remote speech recognition processor, a name corresponding to the feature signal;

retrieving a telephone number corresponding to the retrieved name; and

automatically dialing the number retrieved.

2. The method of claim 1, wherein the step of locally conditioning the audio signal includes the substeps of:

converting the audio signal from an analog signal to digital signal; and

extracting the set of semantic feature characteristics from the digital signal.

3. The method of claim 1, wherein the step of locally forming the feature signal includes the substep of compressing the set of extracted semantic feature characteristics for transmission over the control channel.

4. The method of claim 3, wherein the step of locally compressing includes the substep of:

compressing the set of extracted semantic feature characteristics to a bandwidth under 8 Kbps.

5. The method of claim 3, wherein the step of locally compressing includes the substep of:

compressing the set of extracted semantic feature characteristics to a bandwidth under 16 Kbps.

6. The method of claim 1, wherein the remote speech recognition processor includes a directory of names, represented by a sequence of phonemes, and wherein the step of retrieving the name includes the substep of

comparing the feature signal with the sequence of phonemes to identify at least one of the sequence of phonemes that best matches the feature signal.

7. The method of claim 1, wherein the remote speech recognition processor includes a directory of names, wherein the names in the directory are represented by a sequence of orthographic letters, and wherein the step of retrieving the name includes the substep of

comparing the feature signal with the names in the directory using a phoneme-level representation of the names as an intermediary to identify the phoneme-level representation that best matches the feature signal.

8. A voice-activated dialing method comprising the steps of:

receiving via a control channel, at a remote speech recognition processor, a feature signal representing an audio signal;

comparing the received feature signal with a plurality of semantic feature characteristic sets;

selecting at least one of the plurality of semantic feature characteristic sets that best matches the received feature signal;

retrieving a name corresponding to the selected semantic feature characteristic set;

selecting a telephone number corresponding to the retrieved name; and

automatically dialing the selected telephone number.

9. The method of claim 8, wherein the remote speech recognition processor includes a directory of names, represented by a sequence of phonemes, and wherein the step of retrieving the name includes the substep of

comparing the feature signal with the sequence of phonemes to identify at least one of the sequence of phonemes that best matches the feature signal.

10. The method of claim 8, wherein the remote speech recognition processor includes a directory of names, wherein the names in the directory are represented by a sequence of orthographic letters, and wherein the step of retrieving the name includes the substep of

comparing the received feature signal with the names in the directory using a phoneme-level representation of the names as an intermediary to identify the phoneme-level representation that best matches the feature signal.

11. A method in a voice-activated calling system comprising the steps of:

receiving, at telephone terminal, an audio signal corresponding to a desired destination;

conditioning, locally at the telephone terminal, the audio signal to extract a set of semantic feature characteristics;

forming, locally at the telephone terminal, a feature signal representing the set of semantic feature characteristics corresponding to of the audio signal; and

transmitting the feature signal via a control channel to a remote speech recognition processor.

12. The method of claim 11, wherein the step of conditioning includes the substeps of

converting the audio signal from an analog to digital signal; and

extracting the set of semantic feature characteristics from the digital signal.

13. The method of claim 11, wherein the step of forming the feature signal includes the substep of

compressing the set of extracted semantic feature characteristics for transmission over the control channel.

14. The method of claim 13, wherein the step of compressing includes the substep of

compressing the set of extracted semantic feature characteristics to a bandwidth under 8 Kbps.

15. The method of claim 13, wherein the step of compressing includes the substep of

compressing the set of extracted semantic feature characteristics to a bandwidth under 16 Kbps.

16. A voice-activated calling system comprising:

a telephone terminal for receiving an audio signal corresponding to a destination;

means, coupled to the telephone terminal, for locally conditioning the audio signal to extract a set of semantic feature characteristics;

means, connected to means for locally conditioning, for locally forming a feature signal representing the set of semantic feature characteristics of the received audio signal;

means, connected to the means for locally forming, for transmitting the feature signal via a control channel; and

means, connected to the means for transmitting, for retrieving a name corresponding to a received feature signal.

17. The system of claim 16, wherein means for locally conditioning further includes

means for converting the audio signal to a digital signal; and

means, coupled to means for converting, for extracting a set of feature characteristics from the digital signal.

18. The system of claim 16, wherein the means for locally forming further includes

means for compressing the extracted set of feature characteristics for transmission over the control channel.

19. The system of claim 18, wherein means for compressing includes

means for compressing the set of extracted semantic feature characteristics to a bandwidth under 8 Kbps.

20. The system of claim 18, wherein means for compressing includes

means for compressing the set of extracted semantic feature characteristics to a bandwidth under 16 Kbps.

21. The system in claim 16, further including

means, connected to means for receiving, for storing a plurality of names and corresponding numbers;

means for selecting a number from means for storing, wherein the number selected corresponds to the name retrieved by means for retrieving; and

means for automatically dialing the number selected.

22. The system of claim 16, wherein means for retrieving includes a remote speech processor, and wherein the remote speech processor includes

means, connected to means for transmitting, for receiving the feature signal;

a directory of speech recognition templates, wherein the templates include a sequence of phonemes corresponding to orthographic letters of the names in means for storing; and

means, coupled to the directory, for comparing the received feature signal with the sequence of phonemes to identify at least one of the sequence of phonemes that best matches the feature signal.

23. The system of claim 16, wherein means for retrieving includes a remote speech processor, and wherein the remote speech processor includes

means, connected to means for transmitting, for receiving the feature signal;

a directory of names, wherein the names in the directory are represented by a sequence of tri-phonemes corresponding to orthographic letters of the names in the directory; and

means, coupled to the directory, for comparing the feature signal with the sequence of tri-phonemes to identify at least one of the sequence of tri-phonemes that best matches the feature signal.

24. A telephone terminal apparatus comprising:

means for receiving an audio signal corresponding to a desired destination;

means for locally conditioning the audio signal to extract a set of semantic feature characteristics;

means for locally forming a feature signal corresponding to the extracted set of semantic feature characteristics; and

means for transmitting the feature signal to a remote speech recognition processor via a control channel.

25. The terminal of claim 24, wherein means for locally conditioning further includes:

means for converting the audio signal to a digital signal; and

means for extracting a set of semantic feature characteristics from the digital signal.

26. The system of claim 24, wherein means for locally forming further includes:

means for compressing the extracted set of semantic feature characteristics for transmission over the control channel.

27. The terminal of claim 26, wherein means for compressing includes

means for compressing the set of extracted semantic feature characteristics to a bandwidth under 8 Kbps.

28. The terminal of claim 26, wherein means for compressing includes

means for compressing the set of extracted semantic feature characteristics to a bandwidth under 16 Kbps.