Voice actuation with contextual learning for intelligent machine control

• US 20010047265A1
• Filed: 03/02/2001
• Published: 11/29/2001
• Est. Priority Date: 03/02/2000
• Status: Active Grant

• Current Assignee: Raymond Sepe Jr
• Original Assignee: Raymond Sepe Jr
• Inventors: Sepe, Raymond Jr.
• Granted Patent: US 6,895,380 B2
• Time in Patent Office: Days
• US Class Current: 704/275
• CPC Class Codes: G10L 15/183 using context dependencies, e.g. language models
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/223 Execution procedure of a spoken command
• CPC Class Codes: G10L 2015/227 of the speaker; Human-factor methodology

Abstract

An interactive voice actuated control system for a testing machine such as a tensile testing machine is described. Voice commands are passed through a user-command predictor and integrated with a graphical user interface control panel to allow hands-free operation. The user-command predictor learns operator command patterns on-line and predicts the most likely next action. It assists less experienced operators by recommending the next command, and it adds robustness to the voice command interpreter by verbally asking the operator to repeat unlikely commanded actions. The voice actuated control system applies to industrial machines whose normal operation is characterized by a nonrandom series of commands.

Claims

1. A voice actuated system with contextual learning for intelligent machine control, said system comprising:

voice recognition system receiving voice inputs and identifying one or more voice commands from said received voice input;

command predictor identifying a probability of likeliness of occurrence of said identified one or more voice commands via a statistical likelihood estimation, said command predictor validating said identified one or more voice commands for execution in said machine, and

command processor receiving and executing in said machine said validated one or more voice commands.

2. A voice actuated system with contextual learning for intelligent machine control, as per

claim 1

, wherein said command predictor validates said identified one or more commands for execution in said machine if said probability of likeliness of occurrence is greater that a threshold probability and said command predictor updates said probability of likeliness of occurrence.

3. A voice actuated system with contextual learning for intelligent machine control, as per

claim 2

, said voice recognition system requests for additional voice inputs for clarification if said probability of likeliness of occurrence of said one or more voice commands is below said threshold value, and said command predictor validates said one or more commands for execution in said machine upon receiving said additional voice inputs, and said command processor updates said probability of likeliness of occurrence.

4. A voice actuated system with contextual learning for intelligent machine control, as per

claim 1

, wherein said system further comprises a user interface providing intelligent assistance by revealing a list of probabilities defining the likelihood that one or more commands are next to be executed in said machine.

5. A voice actuated system with contextual learning for intelligent machine control, as per

claim 4

, wherein said user interface includes displaying:

said list of probabilities;

one or more parameters associated with said machine;

one or more graphs illustrating an output associated with said machine, and

one or more commands for controlling said machine.

6. A voice actuated system with contextual learning for intelligent machine control, as per

claim 5

, wherein a command corresponding to the highest probability in said list of probabilities is visually modified to indicate that it is the next likely command.

7. A voice actuated system with contextual learning for intelligent machine control, as per

claim 1

, wherein said command predictor is based on a statistical Markov model.

8. A voice actuated system with contextual learning for intelligent machine control, as per

claim 1

, wherein said machine is an intelligent vehicle.

9. A voice actuated system with contextual learning for intelligent machine control, as per

claim 8

, wherein said intelligent vehicle learns a driver'"'"'s operating patterns and adjusts vehicle handling and performance based on said learned information.

10. A voice actuated system with contextual learning for intelligent machine control, as per

claim 1

, wherein said machine is any of the following:

testing equipment, programmable device or programmable industrial instruments.

11. A voice actuated system with contextual learning for intelligent machine control, as per

claim 1

, wherein said machine is a tensile testing machine.

12. A voice actuated system with contextual learning for intelligent machine control, as per

claim 1

, wherein, in one mode, said command predictor is disabled.

13. A method for voice actuated contextual learning for intelligent machine control, said machine functionally partitioned into one or more discrete states and associating a present condition of said machine with a current state, said method comprising:

receiving one or more voice inputs;

identifying one or more commands from said received voice inputs;

identifying any commands, from said one or more commands, causing transition between said current state and any of said one or more discrete states and identifying probabilities associated with said transitions;

validating one or more commands,

and executing said validated commands in said machine.

14. A method for voice actuated contextual learning for intelligent machine control, said machine functionally partitioned into one or more discrete states and associating a present condition of said machine with a current state, as per

claim 13

, wherein said method further comprises:

checking if each of said identified probabilities is greater than a threshold ‘

t’

, and if so updating corresponding probabilities and validating corresponding commands for execution in said machine, else

requesting another voice input for clarification, and upon clarification, validating corresponding commands, updating corresponding probabilities, and executing said validated one or more commands in said machine.

15. A method for voice actuated contextual learning for intelligent machine control, said machine functionally partitioned into one or more discrete states and associating a present condition of said machine with a current state, as per

claim 13

, wherein said probabilities are Markov chain probabilities.

16. A method for voice actuated contextual learning for intelligent machine control, said machine functionally partitioned into one or more discrete states and associating a present condition of said machine with a current state, as per

claim 13

, wherein said method further comprises displaying a graphical user interface providing intelligent help by displaying a list of probabilities defining the likelihood that said one or more commands are next to be executed in said machine.

17. A method for voice actuated contextual learning for intelligent machine control, said machine functionally partitioned into one or more discrete states and associating a present condition of said machine with a current state, as per

claim 13

, wherein said method is used in an industrial setting to:

reduce operator fatigue, allow freedom of movement, or assist the physically challenged.

18. A method for voice actuated contextual learning for intelligent machine control, said machine functionally partitioned into one or more discrete states and associating a present condition of said machine with a current state, as per

claim 13

, wherein said machine is a tensile testing machine.

19. A graphical user interface for providing intelligent help in a voice actuated system with contextual learning for intelligent machine control, said interface comprising:

a graphical user interface panel displaying various parameters associated with said machine, and

a probabilities panel displaying Markov state of said machine and probabilities associated with one or more commands, said probabilities defining the likelihood that said one or more commands are next to be executed.

20. A graphical user interface for providing intelligent help in a voice actuated system with contextual learning for intelligent machine control, as per

claim 19

, wherein a command corresponding to the highest probability is visually modified to indicate that it is the next likely command.

21. An article of manufacture comprising computer usable medium having computer readable code embodied therein which provides a graphical user interface for providing intelligent help in a voice actuated system with contextual learning for intelligent machine control, said computer readable code comprising:

computer readable program code providing a graphical user interface panel displaying various testing parameters and graphs associated with said machine, and

computer readable program code providing a probabilities panel displaying Markov state of said machine and probabilities associated with one or more commands, said probabilities defining the likelihood that said one or more commands are next to be executed.

22. A voice actuated intelligent machine control system for a tensile testing machine, said system operable in a plurality of modes, said system comprising:

in a first mode,

a voice recognition system receiving voice inputs and identifying one or more voice commands from said received voice input to intelligently control specific parts of said tensile testing machine;

a command predictor identifying a probability of likeliness of occurrence of said identified one or more voice commands via a statistical likelihood estimation, said command predictor validating said identified one or more voice commands for execution in said machine;

in a second mode,

a voice recognition system receiving voice inputs and identifying one or more voice commands from said received voice input to intelligently control specific parts of said tensile testing machine;

a command validator validating said identified one or more voice commands for execution in said machine;

in a third mode,

an input recognition system receiving inputs and identifying one or more commands from said received input to intelligently control specific parts of said tensile testing machine;

a command predictor identifying a probability of likeliness of occurrence of said identified one or more voice commands via a statistical likelihood estimation, said command predictor validating said identified one or more commands for execution in said machine, and

a command processor receiving and executing in said machine said validated one or more commands.

23. A system for intelligent machine control with contextual learning, said system comprising:

interface, said interface receiving inputs and identifying one or more commands from said received inputs;

command predictor identifying a probability of likeliness of occurrence of said identified one or more commands via a statistical likelihood estimation, said command predictor validating said identified one or more commands for execution in said machine, and

command processor receiving and executing in said machine said validated one or more commands.