JPH06131583A - Voice recognition informing system - Google Patents

Abstract

PURPOSE:To automatically inform voice generation when a specific speaker generates specific voice (language). CONSTITUTION:Voice from a voice input part 1 is inputted to a voice analyzing part 2 and its feature is extracted and temporarily recorded in a storage part 3. A judging part 5 compares data read out from the storage part 3 with feature data read out from a memory part 4 and registered through a control part 11, and when both data coincide with each other, a dial signal is generated from a dial signal generating part 7 based upon data read out from a telephone number storing part 6 and outputted from an output part 10 to a telephone line. In response to a terminating call from a communicated subscriber, communication message data are read out from a message storing part 8 and synthesized to a voice by a voice synthesizing part 9 and the voice is sent to the telephone line through the output part 10.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice recognition and reporting system, and more particularly to a system for discriminating a voice of a speaker and notifying a preset addressee.

[0002]

2. Description of the Related Art Some conventional security reporting devices are designed to detect an abnormality with a crime prevention sensor or to automatically report to a required report destination in response to the pressing of an emergency button. When a suspicious person invades a financial institution or store where such a device is installed, if the emergency button can be pressed so that the suspicious person cannot be noticed, a notification is automatically made. In some cases, the operation of the emergency button may be noticed by a suspicious person, so that the operation of the button cannot be performed, and the security device installed at a corner may not be useful.

[0003]

SUMMARY OF THE INVENTION In view of such a point, the present invention registers a voice set in advance for notification of security and the like, compares the uttered voice with the registered voice,
It is an object of the present invention to provide a system for automatically dialing a preset telephone number to call a report destination based on a match and transmitting a security notification message voice in response to an incoming call.

[0004]

In order to solve the above problems, the present invention includes a voice input section for inputting voice, a voice analysis section for analyzing the input voice and extracting characteristic data, and a voice analysis section. Storage unit for storing data of the above, a memory unit for storing voice characteristic data set for reporting when a security abnormality occurs, and comparing the data recorded in the storage unit with the data read out from the memory unit. A discriminating unit that discriminates, a telephone number storage unit that stores a telephone number of a security notification destination, a dial signal generation unit that generates a dial signal based on the data read from the telephone number storage unit, and a message voice such as a security notification. A message storage unit for storing a message, a voice input unit, a voice analysis unit, a storage unit, a memory unit, a determination unit, a telephone number storage unit, a dial signal generation unit, and a message storage unit. The control unit controls the sage storage unit, dials the telephone number read from the telephone number storage unit based on the determination of the determination unit, sends out to the telephone line, and responds to the incoming call from the message storage unit. The present invention provides a voice recognition notification system that outputs the voice of.

[0005]

With the above-described structure, in the voice recognition / report system according to the present invention, the voice input from the voice input unit has its feature data extracted by the voice analysis unit and recorded in the storage unit. The discriminating unit compares the characteristic data with the pre-registered characteristic data read out from the memory unit, discriminates that they are coincident with each other when the coincidence rate is equal to or higher than a predetermined coincidence rate, and pre-registers the telephone number of the notified party based on the discrimination. Read out, dial automatically, and send out via telephone line. The message for notification is read out in response to the incoming call from the callee and is sent to the callee by voice.

[0006]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a voice recognition and notification system according to the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of essential parts of an embodiment of a voice recognition notification system according to the present invention. In the figure, reference numeral 1 is a voice input unit for inputting voice with a microphone or the like. A voice analysis unit 2 analyzes the voice signal from the voice input unit 1 and extracts characteristic data. A storage unit 3 stores the characteristic data analyzed by the voice analysis unit 2. A memory unit 4 stores the set voice feature data. A discriminating unit 5 compares the data read from the storage unit 3 with the data read from the memory unit 4, and outputs a signal when the matching rate is equal to or higher than a predetermined rate. A telephone number storage unit 6 stores the telephone number set as a security notification destination or the like. Reference numeral 7 is a dial signal generator digital, and generates a corresponding dial signal based on the data read from the telephone number memory 6. A message storage unit 8 stores a message set for security notification. A voice synthesizer 9 synthesizes a voice based on the data from the message storage 8. An output unit 10 sends the dial signal from the dial signal generator 7 or the voice from the voice synthesizer 9 to the telephone line. A control unit 11 controls each unit.

Next, the operation of the voice recognition notification system according to the present invention will be described with reference to the flow chart of FIG. When a suspicious person invades, for example, the microphones installed in various places have the preset words "Mr. Sasaki?"
Etc. This word is set as a keyword for security notification, and the voices of "Are you Mr. Sasaki?" To be registered as notification personnel are extracted in advance and registered in the memory unit 4. The signal from the microphone is input from the voice input unit 1 (step 1, hereinafter abbreviated as ST1), the feature is extracted by the voice analysis unit 2 via the control unit 11 (ST2), and the extracted data is once stored in the storage unit 3. Record (ST3).

In response to recording in the storage unit 3, the characteristic data registered in the memory unit 4 is read out via the control unit 11,
The determination unit 5 compares the data read from the storage unit 3 via the control unit 11 (ST4). Then, when the matching rate is equal to or higher than the predetermined match rate (ST5: Yes), the control unit 11 is based on this determination.
(ST6), the telephone number registered as the security notification destination is read from the telephone number storage unit 6, and a desired dial signal is generated through the dial signal voicing unit 7, and the output unit 10 Sent to the telephone line (ST7). Responds to the incoming call response from the other party (ST8: Ye
s), registered in the message storage unit 8, for example,
Message data such as "This is XX of XX-YYYY-ZZZZ, a robber has invaded" is read through the control unit 11, synthesized by the voice synthesis unit 9 into voice, and this voice is output by the output unit 10. It is repeatedly sent to the telephone line via the required number of times (ST9).

If a registered voice that matches the voice from the microphone is not found (ST5: No), all the data registered in the memory unit 4 is sequentially searched, and if none of the data matches (ST10: No). ), It is determined that the voice input from the microphone is not the voice of a person registered as a reporting staff (ST11), and no notification is given as it is not an abnormal situation. The data recorded in the storage unit 3 is erased when the notification device is activated (ST6) or when the discrepancy is determined.

In the above description, the voice of the message is synthesized by the voice synthesizer 9 based on the data from the message storage 8, but the message is recorded on a magnetic recording tape or a compact disc. This voice may be reproduced and transmitted.

[0011]

As described above, according to the voice recognition / report system according to the present invention, when a registered speaker utters a registered voice (a specific word), a security notification is automatically issued. Since it is carried out, for example, if it is installed in a store or the like that is open at night with a small number of store staff, it is convenient that an intruder or the like can be notified without being noticed by an intruder or the like when it is difficult to operate an emergency button or the like.

[Brief description of drawings]

FIG. 1 is a block diagram of a main part of an embodiment of a voice recognition notification system according to the present invention.

FIG. 2 is a flowchart for explaining the operation of the voice recognition notification system according to the present invention.

[Explanation of symbols]

 1 voice input unit 2 voice analysis unit 3 storage unit 4 memory unit 5 discrimination unit 6 telephone number storage unit 7 dial signal generation unit 8 message storage unit 9 voice synthesis unit 10 output unit 11 control unit

Claims (1)

[Claims] 1. A voice input unit for inputting voice, a voice analyzing unit for analyzing the input voice and extracting characteristic data, a storage unit for storing data from the voice analyzing unit, and a notification when a security abnormality occurs. Memory unit for storing voice characteristic data set for use, a judging unit for judging the data recorded in the memory unit by comparing with the data read from the memory unit, and a telephone number of a security report destination, etc. A telephone number storage unit, a dial signal generation unit that generates a dial signal based on the data read from the telephone number storage unit, a message storage unit that stores a message voice such as a security notification, the voice input unit, and a voice analysis unit. Department,
A storage unit, a memory unit, a discrimination unit, a telephone number storage unit, a dial signal generation unit, and a control unit for controlling the message storage unit, and dials the telephone number read from the telephone number storage unit based on the discrimination of the discrimination unit. A voice recognition notification system that makes a call, sends it to a telephone line, and outputs a voice from a message storage unit in response to an incoming call.