JP2839524B2 - Voice input system - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial Application Field) The present invention relates to a voice input system applied to, for example, a system for managing entry permission.

(Prior art) In important facilities such as research laboratories, there are many confidential matters and the like, and it is necessary to regulate the access of people.

The most common method of restricting entry and exit of such people is to place a guard at the entrance. In other words, there is a management room at the entrance, the security guard requests the entrance person to show an ID card, checks the photograph attached to the ID card and the person, and permits entry if the person is the person How to

However, the need for personnel reduction within a company has spread to such administrative departments, and there is a strong demand for unmanned operations by mechanization.

As one example, there is a system in which a keyboard is attached to the entrance and entry is permitted when a beginner inputs a predetermined ID number to the keyboard.

However, such a system has a disadvantage that if another person knows the ID number, it is easy to start. There is also a drawback that if you forget your ID number, you cannot enter.

Another example is a system using a magnetic card. In other words, the system is such that a personal magnetic card is inserted into a card reader attached near the entrance to permit entry of an ID number to be read.

However, this system has a drawback that an introductory person needs time and effort to take out a magnetic card and insert it into a card reader, which is troublesome. Another drawback is that if another person puts his hand in the magnetic card, he can easily get started.

As another example, a system using a wireless card has recently attracted attention. In other words, the beginner is carrying a wireless card, and when approaching the entrance, the ID number is transmitted as a radio wave from the wireless card, and the receiver installed near the entrance receives and decodes this ID number and permits entry. System.

However, this system eliminates the need for an introductory person to take out the card and insert it into a card reader as in the magnetic card system described above, but it is easy to get started when another person obtains and uses the wireless card. There is a similar drawback that you can.

Another interesting example is a speaker verification device. That is, a secret word spoken by an introductory person for each individual (for example, "Hirake Sesame") is registered in advance as an audio pattern, and when the introductory person speaks a secret word toward the machine, the entered secret word is input. Perform pattern matching between the voice pattern and the registered voice pattern,
If the degree of identification exceeds a predetermined reference value, the person is regarded as a person.

However, this speaker verification device has the advantage that an introductory person can identify himself / her with simple operation without directly touching the device in order to utter a voice, but since the voice changes over time, a card or the like is used. It is inferior in reliability compared to the method of identifying the person. In addition, there is a disadvantage that pattern matching becomes difficult if the surrounding noise is large when uttering a voice.

The drawback that this pattern matching becomes difficult is
The distance between the microphone used to input voice and the beginner's mouth becomes a problem. Therefore, when considering the influence of ambient noise, the beginner needs to bring his mouth closer to the microphone within a predetermined distance, which is troublesome for the beginner.

(Problems to be Solved by the Invention) As described above, in the conventional system for managing entry, there are problems in terms of operability, reliability, or reliability.

The present invention has been made to solve such a problem, and an object of the present invention is to provide a voice input system having high operability, reliability, or reliability.

[Structure of the Invention] (Means for Solving the Problems) In order to solve the above problems, the present invention provides a signal related to a predetermined ID number as a response signal when a transmission invitation signal is received in a predetermined communication area. And a receiving means for receiving a signal transmitted from the wireless card,
A storage unit that stores an ID number and a code related to a voice in advance in association with each other, and an ID number included in a signal received by the reception unit and the storage unit that stores the ID number.
Collating means for collating with an ID number, and first collating means for displaying a message indicating that there is a coincidence with the received ID number as a result of collation by the collating means, When there is a match by the collation by the collation unit, a voice input is accepted,
The voice input part is provided at an optimum position of a voice input having a relative positional relationship with the communication area where the transmission invitation signal can be received, and the voice input means is provided to the voice input means. Comparing means for comparing the code relating to the voice with the code relating to the voice stored in the storage means and corresponding to the comparison result; and, as a result of the comparison by the comparing means, whether or not the codes relating to the respective voices match. Is displayed by a message, and a main body including a control means for issuing a predetermined control signal when a result of the match is obtained by the comparing means.

(Operation) In the present invention, a transmission invitation signal is received by a wireless card in a predetermined communication area, and a predetermined ID is transmitted from the wireless card.
When the signal related to the number is transmitted, the main body receives this signal, compares the received ID number with the previously stored ID number, and if there is a match, a message that there is a match is provided by a message. In addition to displaying, for example, a code related to a voice input via a microphone or the like at an optimal position in voice input by a wireless card owner and a code related to a voice corresponding to an ID number stored in advance are compared. Then, as a result of the comparison, whether or not the codes relating to the voices match is further displayed as a message, and a predetermined control signal is transmitted only when the result of the matching is obtained.

Therefore, a voice input system with high operability, reliability, and certainty is realized.

Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing a configuration of a voice input system according to an embodiment of the present invention, and FIG. 2 is an external view thereof.

In FIG. 1, reference numeral 1 denotes a wireless card carried by a beginner, and transmits a signal of an ID number registered in advance in response to a transmission invitation signal from the main body.

On the other hand, reference numeral 2 denotes a transceiver for forming a wireless channel with the wireless card 1 via the antenna 3 and exchanging information. Also,
Reference numeral 4 denotes a microphone for converting a voice uttered by the beginner into an analog signal, and reference numeral 5 denotes a display for instructing the beginner to invite a voice. Further, 6 is a personal voice pattern for each beginner (sent from a voice pattern file 10 described later).
Is a speaker verification device that compares the content of the voice pattern memory 7 for storing a voice signal with the voice signal from the microphone 4 to verify whether or not the uttered voice is authentic. Reference numeral 8 denotes a push button switch used to start registration, 9 denotes a control computer for controlling each unit, 10 denotes a voice pattern file storing a registration pattern for each beginner, and 11 denotes a door of an important facility (FIG. 2). (See Reference).

Next, the operation of the thus configured voice input system of the present embodiment will be described.

First, a beginner carries the wireless card 1 and approaches the antenna 3.

Here, the wireless card 1 normally operates only in the reception function, and the other functions (for example, the transmission function) are in a standby mode in which the wireless card 1 is stopped, so as to prevent battery consumption as much as possible.

On the other hand, the transmitting / receiving signal of the ID number is transmitted from the transceiver 2 through the antenna 3 at regular time intervals. In addition, the reachable distance of the invitation signal is set such that the position where the beginner utters a voice is the optimum distance for the microphone 4 when the microphone 4 faces the beginner (wireless card 1). I have.

Therefore, when the beginner approaches a distance other than the distance set in this way, the transmission invitation signal is received by the wireless card 1.

When the wireless card 1 receives the transmission invitation signal, the mode changes from the standby mode to the normal mode, and the wireless card 1 encodes an ID number uniquely set for the wireless card and transmits the code on a radio wave.

This radio wave is received by the antenna 3, coded by the transceiver 2, and sent to the control computer 9.

The control computer 9 searches the voice pattern file 10 from the transmitted code. Then, a voice pattern corresponding to the code is extracted and stored in the registered pattern memory 7 through the speaker verification device 6.

Then, the control computer 9 controls the display 5 to display the following.

“Please say the secret word.” The control computer 9 sends a command to start the verification process to the speaker verification device 6.

The speaker verification device 6 starts inputting voice from the microphone 4.

Here, the beginner stands at an optimal position with respect to the front of the microphone. This is, as described above, such that, when the microphone 4 and the beginner (wireless card 1) face each other, the position at which the beginner utters a voice is optimal for the microphone 4 as described above. Because it is set.

Then utter the password. For example, "Hirake Sesame"
Or it may be your own name.

The speaker verification device 6 performs a verification process between the voice input through the microphone 4 and the voice pattern stored in the registered pattern memory 7. In general, as a simple method of this collation processing, there is a DP matching method or the like. Then, a score is calculated in the course of the collation processing, and if this score is larger than a predetermined numerical value, the user is identified and the result is sent to the control computer 9.

The control computer 9 controls the display 5 to perform, for example, the following display.

“The entry is permitted.” At the same time, the control computer 9 sends a control signal to the door opening / closing mechanism 11. As a result, the door opening / closing mechanism 11 opens the door 12. After a certain time, the door opening / closing mechanism 11 closes the door 12, and a series of sequences ends.

On the other hand, if the score is low in the verification process in the speaker verification device 6, the control computer 9 controls the display 5 to display, for example, the following.

"I couldn't identify myself. I can't get started." After a certain period of time, the display disappears and the sequence ends.

Thus, according to the system of the present embodiment, the wireless card 1
Is set such that when the microphone 4 and the beginner (wireless card 1) face each other, the position at which the beginner utters a voice is the optimum distance for the microphone 4. Therefore, ambient noise is less likely to be mixed in, and optimal sound processing can be performed.

 Next, the registration operation will be described.

As described above, when performing speaker verification by voice, it is necessary to consider that the voice of the beginner may change over time. Therefore, beginners need to periodically re-register the voice pattern.

In this system, when registration is desired, the wireless card 1 of the beginner is carried and stands in front of the microphone 4.
Here, the sequence in which the main body reads the ID number from the wireless card 1 is the same as in the case of the above-described collation. The difference is that the beginner then presses the registration switch 8.

When detecting that the registration switch 8 has been pressed, the control computer 9 controls the display 5 to display, for example, the following.

“Please say a secret word for registration.” At the same time, the control computer 9 outputs a command to start the registration to the speaker verification device 6. In addition, microphone 4
Starts voice input.

When the beginner utters a password, the voice is stored once in the registered pattern memory 7. When the utterance of the voice ends, the voice is read out by the control computer 9 and stored in the voice pattern file 10 in association with the ID number.

Thereafter, the control computer 9 controls the display 5 again to display, for example, the following display.

 "Registration is complete." The registration sequence is now completed.

[Effects of the Invention] As described above, according to the present invention, when a signal related to a predetermined ID number is transmitted from a wireless card, the main body receives this signal, and the received ID number is stored in advance. ID
If there is a match with the number, a message indicating that there is a match is displayed with a message, and the code related to the voice input from the wireless card owner and the voice corresponding to the ID number stored in advance are displayed. This code is compared with the relevant code, and as a result of the comparison, whether or not the code related to the voice is matched is further displayed by a message, and a predetermined control signal is issued only when the result of the match is obtained, so that the operability is improved. Thus, a highly reliable or reliable voice input system is realized.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a voice input system according to an embodiment of the present invention, and FIG. 2 is an external view thereof. 1 ... wireless card, 2 ... transceiver, 3 ... antenna,
4 microphone, 5 display, 6 speaker verification device, 7 registered pattern memory, 8 registration switch,
9: Control computer, 10: Voice pattern file, 11
…… door opening and closing mechanism, 12 …… door.

Claims (1)

(57) [Claims] A wireless card for transmitting a signal corresponding to a predetermined ID number as a response signal when a transmission invitation signal is received in a predetermined call area; and a receiving means for receiving a signal transmitted from the wireless card. A storage unit that stores an ID number and a code related to voice in advance, and compares an ID number included in a signal received by the reception unit with an ID number stored by the storage unit. A first display means for displaying a message indicating that there is a match with the received ID number as a result of the check by the check means, and a check by the check means. When there is a match, the voice input is accepted, and the input portion of the voice has a relative positional relationship with the communication area where the transmission invitation signal can be received. Voice input means provided at an optimal position, and a comparison means for comparing a code relating to the voice input to the voice input means with a code relating to the voice stored in the storage means and corresponding to the comparison result. A second display means for displaying as a message whether or not the codes related to the respective voices match as a result of the comparison by the comparing means, and a predetermined control signal when the matching result is obtained by the comparing means. A voice input system comprising: a main body having control means for emitting the voice.