Method and apparatus for including speech recognition capabilities in a computer system

• US 5,659,665 A
• Filed: 12/08/1994
• Issued: 08/19/1997
• Est. Priority Date: 12/08/1994
• Status: Expired due to Fees

• Current Assignee: Lucent Technologies, Inc. (Nokia Corporation)
• Original Assignee: Lucent Technologies, Inc. (Nokia Corporation)
• Inventors: Whelpley, James Hall Jr.
• Primary Examiner(s): MacDonald, Allen R.
• Assistant Examiner(s): EDOUARD, PATRICK NESTOR
• Application Number:
  US08/352,251
• Time in Patent Office: 985 Days
• Field of Search: 395/2.84, 395/2, 395/2.79, 395/2.4, 395/2.55, 381/42-45
• US Class Current: 704/275
• CPC Class Codes: G06F 3/16 Sound input; Sound output speech processing G10L
• CPC Class Codes: G10L 15/28 Constructional details of speech recognition systems

Abstract

A computer system retrofit is described which allows the inclusion of voice recognition capabilities to provide a user of a standard system not incorporating voice recognition with the ability to use voice commands when editing a document on a computer system. An apparatus that recognizes speech patterns may be inserted in a computer system as a stand alone unit that provides keystroke data to the standard keyboard cable input port of a microprocessor unit. The apparatus, based on the detection of predetermined voice commands, generates keystroke data which are equivalent to the keystroke data which would have been generated had the computer user physically depressed a corresponding sequence of keys at a keyboard.

Claims

1. A computer system that generates keystroke data from the detection of speech patterns, comprising:

a keyboard for supplying keyboard keystroke data on a keyboard cable attached to the keyboard;

a microprocessor unit having a standard keyboard cable input port connector;

a low cost voice command device in a self-contained housing having a limited voice command recognition capability, to reduce the necessity for frequent execution of repeated keystroke sequences at the keyboard by a computer user, said low cost voice command device having a keystroke data input and output port, a memory for storing a predetermined number of plain English voice command words associated with the performance of repetitive multiple keystroke sequences said voice command words comprising cursor movement control words and data information entry word, a receiver for detecting speech patterns and a processing component for generating voice keystroke data, wherein said keyboard cable connects said keyboard to said input port of said voice command device;

a standard computer cable connecting said output port of said voice command device to said standard keyboard cable input port of said microprocessor unit,

wherein said voice command device receives keyboard keystroke data at the input port and the processing component generates voice keystroke data based on the recognition of the speech patterns of a voice command word which is provided by a user and detected at the receiver, said voice command word being associated with the performance of a repetitive keystroke sequence on the keyboard,

wherein the voice keystroke data generated are equivalent to the keyboard keystroke data which would have been generated at the keyboard had the user depressed a corresponding sequence of keys on the keyboard,

and wherein said voice command device provides an interleaved voice and keyboard keystroke data stream to the output port for routing to the microprocessor unit.

2. The system of claim 1, wherein the voice command device obtains electrical power for operation which is transmitted over the cable connecting the voice command device to the keyboard cable input port of the microprocessor unit.

3. The system of claim 1, wherein the voice command device further comprises a switch for activating the voice recognition operating feature.

4. The system of claim 1, wherein the voice command device further comprises an indicator light for indicating to the user that voice keystroke data have not been provided to the microprocessor unit when speech patterns are detected at the receiver.

5. The system of claim 1, wherein the recognized voice command comprises a number word and a command word, and wherein the voice command device generates voice keystroke data which are equivalent to the keyboard keystroke data which would have been generated by the physical depression of a key on the keyboard, corresponding to the command movement word, the same number of times as indicated by the value of the number word.

6. The system of claim 5, wherein the command word is a cursor movement word.

7. The system of claim 5, wherein the command word is a data entry word.

8. The system of claim 1, wherein voice keystroke data are included in the interleaved stream only when keyboard keystroke data are not being received at the voice command device, such that keyboard keystroke data are provided in the interleaved stream with absolute priority over voice keystroke data associated with voice commands.

9. The system of claim 5, wherein voice keystroke data are included in the interleaved stream only when keyboard keystroke data are not being received at the voice command device, such that keyboard keystroke data are provided in the interleaved stream with absolute priority over voice keystroke data associated with voice commands.

10. The system of claim 1, wherein keyboard keystroke data which is received at the voice command device during voice command detection or processing is included in the interleaved stream after voice keystroke data generated based on the detected or processed voice command is provided for routing to the microprocessor unit in the interleaved stream.

11. A low cost apparatus comprising:

a first input for receiving keyboard keystroke data generated by a keyboard;

a receiver for detecting speech information;

a processing component for determining whether the speech pattern of any detected speech information corresponds to that of a plain English command word included in a limited number of predetermined voice command;

a memory for storing said limited number of predetermined voice commands, said predetermined voice commands comprising cursor movement control words and data information entry words;

wherein the command word is associated with the performance of a repetitive multiple keystroke sequence on the keyboard, and

wherein said processing component generates voice keystroke data based on the recognition of a voice command and combines voice keystroke data with keyboard keystroke data to provide an interleaved keystroke data stream; and

an output port upon which the interleaved keystroke data stream may be supplied for routing to a keyboard cable input port connector of a microprocessor unit in a computer system.

12. The apparatus of claim 11, wherein the voice command comprises a number word and a cursor movement word, and wherein the processing component generates voice keystroke data which are equivalent to the keyboard keystroke data which would have been generated by the physical depression of a key on the keyboard, corresponding to the cursor movement word, the same number of times as indicated by the value of the number word.

13. The apparatus of claim 11, wherein the processing component processes and detects speech information for inclusion as voice keystroke data in the interleaved stream only when keyboard keystroke data are not being received at the first input, such that keyboard keystroke data are provided in the interleaved stream with absolute priority over voice keystroke data associated with voice commands.

14. The apparatus of claim 11, wherein the receiver comprises a microphone.

15. The apparatus of claim 11 further comprising a switch for controlling voice recognition by the apparatus.

16. The apparatus of claim 11 further comprising an indicator light for indicating that detected speech patterns have not been provided as converted voice keystroke data to the output port.

17. A method of using speech recognition techniques for generating keystroke data based on the detection of a limited set of voice commands comprising cursor movement control words and data information entry words instead of the execution of a repetitive multistroke keystroke sequence on a keyboard, comprising the steps of:

detecting the speech pattern of plain English speech information which is voiced by a computer user;

determining whether the detected speech pattern matches a pattern of a predetermined word which is included in a voice command from the limited set of voice commands; and

,

generating keystroke data representative of the words in the voice command whose speech pattern is recognized, wherein the generated keystroke data are equivalent to the multistroke keystroke data which a computer user would have generated at the keyboard had the user depressed a corresponding sequence of keys on the keyboard.

18. The method of claim 17, wherein the voice command comprises a number word and a cursor movement word, and wherein the step of generating keystroke data further comprises the steps of generating voice keystroke data which are equivalent to the keystroke data which would have been generated by the physical depression of a key on the keyboard, corresponding to the cursor movement word, the same number of times as indicated by the value of the number word.

19. The method of claim 17, further comprising the step of providing a stream of interleaved keystroke data comprised of keystroke data generated based on the recognized voice commands and the keyboard keystroke data which may be supplied from the keyboard.

20. The method of claim 19, further comprising the step of including voice keystroke data in the interleaved stream keystroke data only when keyboard keystroke data is not supplied from the keyboard, such that keyboard keystroke data are provided in the interleaved stream with absolute priority over voice keystroke data associated with voice commands.

21. A computer system that generates keystroke data from the detection of speech patterns, comprising;

a microprocessor unit having a standard keyboard cable input port connector;

a keyboard for supplying keyboard and voice keystroke data on a keyboard cable attached to the keyboard, said keyboard comprising a receiver for detecting speech patterns and a processing component for generating voice keystroke data,

wherein said keyboard cable connects said keyboard to said standard keyboard cable input port of said microprocessor unit,

wherein said keyboard comprises key inputs for generating keyboard keystroke data,

wherein said processing component generated voice keystroke data based on the recognition of speech patters of a plain English voice command word which is provided by a user and detected at the receiver, said voice word command being associated with the performance of a repetitive multistroke keystroke sequence on the keyboard and being selected from a group of cursor control words comprising the words "top", "space", "tab", "page", "right", "left", "up" and "down", from a group of data information entry words "enter", "tab", or "delete," or a number word,

wherein the voice keystroke data generated are equivalent to the keyboard keystroke data which would have been generated at the keyboard had the user depressed a corresponding sequence of keys on the keyboard,

and wherein said keyboard provides an interleaved voice and keyboard keystroke data stream to the keyboard cable for routing to the microprocessor unit.

22. The system of claim 21, wherein the keyboard obtains electrical power for operation from the keyboard cable connecting the keyboard to the keyboard cable input port of the microprocessor unit.

23. The system of claim 21, wherein the keyboard further comprises a switch for activating the voice recognition operating feature.

24. The system of claim 21, wherein the keyboard further comprises an indicator light for indicating to the user that voice keystroke data have not been provided as converted to the microprocessor unit when speech patterns are detected at the receiver.

25. The system of claim 21, wherein the recognized voice command comprises a number word and a command word, and wherein the keyboard generates voice keystroke data which are equivalent to the keyboard keystroke data which would have been generated by the physical depression of a key input on the keyboard, corresponding to the command movement word, the same number of times as indicated by the value of the number word.

26. The system of claim 21, wherein the command word is a cursor movement word.

27. The system of claim 21, wherein the command word is a data entry word.