Method and system for providing security using voice recognition

• US 6,263,311 B1
• Filed: 01/11/1999
• Issued: 07/17/2001
• Est. Priority Date: 01/11/1999
• Status: Expired due to Term

• Current Assignee: Hanger Solutions, LLC (IP Investments Group LLC)
• Original Assignee: Advanced Micro Devices, Inc.
• Inventors: Dildy, Robert
• Primary Examiner(s): Dorvil, Richemond
• Application Number:
  US09/228,475
• Time in Patent Office: 918 Days
• Field of Search: 704/270, 704/275, 704/273, 704/272, 704/246, 704/247, 704/250, 704/252, 704/231, 704/236, 704/200
• US Class Current: 704/273
• CPC Class Codes: G10L 17/00 Speaker identification or verification techniques

Abstract

Method and system for providing security to a space by detecting an unauthorized presence in the space using voice recognition. A plurality of acceptable voice patterns is generated from sounds of authorized users and stored in the security system. Sounds in the space to be protected are detected and a current voice pattern is generated from at least one sound detected in the space to be protected. The current voice pattern is compared with entries in the stored voice lists to determine a level of difference or a level of similarity between the current voice pattern and the content of the voice list. If the determined level constitutes a discrepancy according to predetermined criteria, an alarm response is initiated. Further, if the determined level constitutes a matching between the current voice pattern and an entry in an emergency list of voice codes that were entered by authorized users, a predetermined emergency action is performed. The advantages of the voice recognition as a security measure may be also integrated into existing security systems.

Claims

1. A method for determining an unauthorized presence in a space to be protected comprising:

storing a voice list, wherein said voice list includes a plurality of voice patterns;

detecting a sound in said space, wherein said sound is indicative of a presence of at least one source of said sound in said space;

generating a current voice pattern from said sound;

comparing said current voice pattern with at least one of said plurality of voice patterns in said voice list to determined if said presence in said space is said unauthorized presence; and

initiating an alarm response if said presence in said space is said unauthorized presences;

wherein said plurality of voice patterns are generated from sounds of one or more authorized users and sounds from one or more non-human sources.

2. The method as recited in claim 1 wherein said generating said current voice pattern from said sound includes using voice recognition to analyze said sound.

3. The method as recited in claim 2 wherein said comparing includes using said voice recognition to determine a first level of discrepancy between said current voice pattern and a content of said voice list, wherein said first level of discrepancy is determined based on first predetermined criteria.

4. The method as recited in claim 3 wherein said first level of discrepancy is used to determined if said presence in said space is said unauthorized presence based on said first predetermined criteria.

5. The method as recited in claim 1 wherein said alarm response includes at least one action of a plurality of actions, wherein said action includes activating at least one device from a plurality of devices, and wherein if at least two actions of said plurality of actions are executed, said at least two actions are executed in a first sequence from a plurality of sequences based on second predetermined criteria.

6. The method as recited in claim 5, further comprising storing an emergency voice code list, wherein said emergency voice code list includes a plurality of voice codes, wherein said plurality of voice codes are generated by said sounds of said one or more authorized users and sounds from said one or more non-human sources, wherein if said current voice pattern corresponds to one of said plurality of voice codes in said emergency voice code list, at least one predetermined action of said plurality of actions is initiated.

7. A system for providing security to a space by detecting an unauthorized presence in said space, the system comprising:

a first sensor configured to detect a sound in said space, wherein said sound is indicative of at least one source of said sound in said space, and wherein said source is indicative of a presence in said space;

a voice-processing unit coupled to said first sensor and configured to store a voice list, wherein said voice list includes a plurality of voice patterns, analyze said sound in said space, generate a current voice pattern from said sound detected by said first sensor, compare said current voice pattern with at least one of said plurality of voice patterns in said voice list to determined if said presence in said space is said unauthorized presence; and

a control unit coupled to said voice-processing unit and said first sensor and configured to initiate an alarm response if said presence in said space is said unauthorized presence;

wherein said plurality of voice patterns are generated from sounds of one or more authorized users and sounds from one or more non-human sources.

8. The system as recited in claim 7 further comprises a first control device from a plurality of devices coupled to said control unit, wherein said first control device is used to perform a first control function.

9. The system as recited in claim 8 wherein said alarm response initiated by said control unit includes at least one action of a plurality of actions, wherein said action includes activating said at least one of said plurality of devices.

10. The system as recited in claim 9 wherein said control unit is further configured to execute at least two actions of said plurality of actions in a first sequence from a plurality of sequences based on second predetermined criteria.

11. The system as recited in claim 9 wherein said voice-processing unit is further configured to store an emergency voice code list, wherein said emergency voice code list includes a plurality of voice codes, wherein said plurality of voice codes are generated by said sounds of said one or more authorized users and sounds from said one or more non-human sources.

12. The system as recited in claim 11 wherein if said voice-processing unit determines that said current voice pattern corresponds to one of said plurality of voice codes in said emergency voice code list, said control unit initiates at least one predetermined action from said plurality of actions.

13. The system as recited in claim 7 wherein said voice-processing unit generates said plurality of voice patterns by recording said sounds of said one or more authorized users and sounds from said one or more non-human sources.

14. The system as recited in claim 7 wherein said voice-processing unit uses voice recognition to recognize said sound, generate said current voice pattern from said sound, and compare said current voice pattern with a content of said voice list.

15. The system as recited in claim 14 wherein said voice-processing unit comparing said current voice pattern with at least one of said plurality of voice patterns in said voice list to determine if said presence in said space is said unauthorized presence includes using a first level of discrepancy between said current voice pattern and said content of said voice list, wherein said first level of discrepancy is based on first predetermined criteria, and wherein said first level of discrepancy is used to determine if said presence in said space is said unauthorized presence based on said first predetermined criteria.

16. The system as recited in claim 7, further comprising a second sensor configured to detect a parameter other than said sound, wherein said second sensor is coupled to said control unit, wherein said control unit is further configured to use an output of said second sensor to initiate a control action based on third predetermined criteria.