JPH11249683A - Voice input device - Google Patents

Abstract

(57) [Summary] [Problem] Even if a password is input by voice in a situation where it can be heard by another person, the input is not specified by the other person and the user is not even aware that the password is being input. Is provided. SOLUTION: A voice input device recognizes a series of input voices input for inputting a password, converts the input voice into voice data, and outputs the voice data. The voice input device includes one or more specific words serving as a password. Specific word storage means 2 to be stored in advance
1 and whether or not the voice data output from the voice recognition input means 1 includes a specific word stored in the specific word storage means 21. And a password identification means 2 for identifying whether or not a password is included in the password.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice input device, and more particularly to a voice input device in which a password is input by voice input.

[0002]

2. Description of the Related Art Conventional password input is performed, for example, in Japanese Patent Application Laid-Open No. 9-213061 or Japanese Patent Application Laid-Open No. 63-223794.
Japanese Patent Laid-Open No. 7-3201 discloses a method of inputting a password using a numeric keypad or a keyboard.
There is a method of selecting and inputting a password element from those displayed as disclosed in Japanese Patent Publication No. 24.

Further, as disclosed in Japanese Patent Application Laid-Open Nos. 8-84190 and 8-223281, input voices are analyzed, voiceprint matches are determined, and individuals are identified to identify devices. In some cases, use permission is given, and in some cases, the user is forced to utter a specific word in order to make it easier to detect coincidence of voiceprint data.

[0004] As a practical application of the conventional speech recognition technology, for example, Vo Japan of IBM Japan, Ltd.
There is ice Type (product name). This is a speech-to-text conversion system, and does not include a password function.

[0005]

For example, when input means for a micro cellular phone is considered, it becomes difficult to input information with a human finger. Therefore, it is necessary to input information (including dial number input) to the telephone by voice using a voice input device, and there may be no means other than voice input. On the other hand, mobile phones
Due to its nature, it may be lost, and restricting its use by inputting a password or the like is an effective measure against plagiarism.

A technique for recognizing a user by voice input is conceivable in order to eliminate troublesome operations using a keyboard or the like. The above-mentioned JP-A-8-84190 and JP-A-8-84190
The example of JP-A-223281 is this example, but in this example, the individual user is specified by voiceprint analysis, and it is difficult to always utter with the same voice quality due to the physical condition and mood of the user at that time. There is a problem in that it is impossible to exclude the possibility that a person other than a regular user uses the same due to a certain voice quality or “voice imitation”.

Therefore, a method has been considered in which a specific word is stored as a keyword by a user, and the keyword is uttered by the user to perform recognition by a voice recognition technique.

[0008] However, when the password is input by using a voice without using a keypad or the like, the password input voice can be heard by others around the user, and there is a problem from the viewpoint of keeping the password itself confidential. Remains.

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and has as its object to input a password by voice in a situation where it can be heard by another person without specifying the password by another person. Another object of the present invention is to provide a voice input device capable of realizing an input in which a user is not even aware that a password has been input.

[0010]

According to a first aspect of the present invention, there is provided a voice input device for recognizing a series of input voices input for inputting a password and converting the input voices into voice data. Voice recognition input means for outputting, a specific word storage means for storing one or more specific words serving as a password in advance, and, among the voice data output from the voice recognition input means, the specific word storage means It is characterized in that it is provided with a password identification means for detecting whether or not the stored specific word is included, thereby identifying whether or not the password is included in the series of input voices.

According to the above arrangement, even if the password is input by voice in a situation where it can be heard by another person, the password is not specified by the other person, and the user is not even aware that the password is being input. Input can be realized.

That is, when it is detected that a specific word is included in a series of input voices, a password input method is adopted which assumes that password input has been completed. I can't tell if I'm warm, and I don't even notice that I'm entering my password.

According to a second aspect of the present invention, there is provided a voice input device according to the first aspect, wherein:
It is characterized by comprising a password identification activation / termination means for activating and / or ending the operation of the password identification means.

According to the above configuration, it is possible to make it more difficult for the observer to determine what the password is. When a password is notified by a password discriminating activation / termination means at the time of password entry, a password input range is notified to the machine out of a series of input voices. Does not know at which point the word was the password.

According to a third aspect of the present invention, there is provided a voice input device according to the first or second aspect, wherein a series of input voices or voice data is recorded and the voice or voice data is recorded. Recording and reproducing means for reproducing the password, wherein the password identifying means identifies the password based on the reproduced voice or voice data.

According to the above configuration, it is possible to shift the time of inputting the password and the time of identifying the password, so that it is more difficult to determine what the password is.

According to a fourth aspect of the present invention, there is provided a voice input device according to the third aspect, wherein:
During playback, the password identification activation / termination means
It is characterized in that the password identification means is activated and / or terminated.

According to the above configuration, it becomes more difficult for the observer to decipher the password. When the user informs the machine of the password input status and starts up during playback of a sound including a specific word, the machine can recognize the word as the entered password, but the observer cannot hear the playback sound. Therefore, even when observing the operation of informing the machine of the user's password input status, it is not possible to identify which voice is the password.

According to a fifth aspect of the present invention, there is provided a voice input device according to the third or fourth aspect, wherein:
It is characterized in that the time from recording to the start of reproduction is fixed in advance or can be set arbitrarily.

According to the above configuration, the operation is simplified if the time is fixed in advance. Also, by changing the time from when the input and recorded sound starts to be played back, even if the observer records the password input sound, measures the time from the end of the input, and compares it with the recording, the password can be changed. Cannot be identified.

According to a sixth aspect of the present invention, there is provided a voice input device according to any one of the first to fifth aspects, wherein a plurality of specific words are stored in the specific word storage means. The password identification unit identifies that the password has been correctly input by detecting that some or all of the plurality of specific words are included in the voice data output from the voice recognition input unit. It is characterized by:

According to the above configuration, a password can be input by having a plurality of password target words and matching a part or all of the words, so that a specific word can be specified no matter how many times it is observed. Can be difficult.

According to a seventh aspect of the present invention, there is provided a voice input device according to the sixth aspect, wherein:
According to a feature of the present invention, the contents accessible after the password is input differ depending on the type of the specific word included in the voice data detected by the password identification unit.

According to the above configuration, the subsequent operation can be changed depending on which specific word is used.

According to an eighth aspect of the present invention, there is provided a voice input device according to the sixth aspect, wherein:
A specific word discrimination rule storage unit that specifies an input order of a plurality of passwords is provided, and the password identification unit identifies that a password has been correctly input only when input voice data has the specified order relationship. It is characterized by.

According to the above configuration, the password input is completed only when there is a relationship specified in the input order of a plurality of passwords, thereby making it more difficult for the observer to specify the password.

According to a ninth aspect of the present invention, there is provided a voice input device according to the second aspect, wherein
The present invention is characterized in that the password identification activation / termination means is provided with activation / termination specific word storage means for activating and / or terminating it in accordance with a voice input.

According to the above configuration, if the password identification means adopts an input method activated by voice input, even if the password is uttered before activation, the password is not accepted, and the observer can be confused.

[0029] According to a tenth aspect of the present invention, there is provided a voice input device comprising:
In order to solve the above-mentioned problem, in the configuration according to the second aspect, the password identification start / end unit activates the password identification unit, measures time, and terminates the password identification operation after a lapse of a predetermined time. It is characterized by having a timing means.

According to the above configuration, it is possible to adopt a password input method having an allowable time for inputting a password, and it is possible to prevent an observer from repeatedly inputting a password.

The voice input device according to the invention of claim 11 is:
In order to solve the above-mentioned problem, in the configuration according to claim 2, after the password identification means has been activated by the password identification activation / termination means, the number of input words is counted. And a counting means for terminating the identification operation.

According to the above configuration, a password input method having an allowable number of words for password input can be adopted.
It is possible to prevent an observer from repeatedly inputting a password.

According to a twelfth aspect of the present invention, there is provided a voice input device comprising:
In order to solve the above-mentioned problem, in the configuration according to any one of claims 1 to 11, a specification in which contents of the specific word storage unit and / or the start / end specific word storage unit is changed by input or reproduced voice data. It is characterized by having word updating means.

According to the above configuration, since the specific word to be stored can be changed, the specification of the password can be made more difficult.

According to a thirteenth aspect of the present invention, there is provided a voice input device comprising:
In order to solve the above-mentioned problem, in the configuration according to any one of claims 1 to 11, the content of the specific word storage unit is input by a word input immediately after or a few words after a specific word to be a password. And a specific word updating means for changing

According to the above configuration, when the password is input, the word pronounced after the password becomes the next password, and the password is changed every time, so that it becomes more difficult for the observer to specify the password.

According to a fourteenth aspect of the present invention, there is provided a voice input device comprising:
In order to solve the above-mentioned problem, any one of claims 1 to 13 is characterized in that a password input success / failure display means for displaying success / failure of a password input after a series of voice inputs is completed.

According to the above configuration, the display allows the operator to know the success or failure of the password input.

A speech input device according to a fifteenth aspect of the present invention
In order to solve the above-mentioned problem, the configuration according to any one of claims 1 to 14, wherein the contents of the specific word storage unit and / or the activation / end specific word storage unit are initialized. .

According to the above configuration, the user can use the apparatus from the time of acquisition by inputting a specific word that is initially set by voice, thereby improving user convenience.

A speech input device according to a sixteenth aspect of the present invention
In order to solve the above problem, a series of input voices input for password input is recognized, and voice recognition input means for converting the input voice into voice data and outputting the voice data, and one or more specific words serving as a password are specified in advance. A specific word storage unit for storing, and detecting whether or not the specific word stored in the specific word storage unit is included in the voice data output from the voice recognition input unit; A password identification unit that detects whether there is a voice input just before or after, or before and after, and thereby determines whether the password input is successful.

According to the above configuration, even if the password is input by voice in a situation where it can be heard by another person, the password is not specified by the other person, and the user is not even aware that the password is being input. Input can be realized.

That is, it is detected whether or not a specific word is included in a series of input voices, and whether or not there is a voice input before and after the specific word is detected. , The observer cannot identify which word was the password, and does not even notice that the password has been entered.

A speech input device according to a seventeenth aspect of the present invention
In order to solve the above-mentioned problem, the configuration according to claim 16, further comprising a voice input time measuring unit that measures a time interval when a specific word and a word input before and / or after the specific word are input. The password identification unit performs the determination by including the time interval in a determination element.

According to the above configuration, since the judgment is performed by including the time interval in the judgment element, it becomes more difficult for the observer to specify the password.

The voice input device according to the invention of claim 18 is
In order to solve the above-mentioned problem, the configuration according to claim 16 further includes a voice input time measuring unit that measures a total time of inputting a specific word and a voice input before and / or after the specific word. The password identification means performs the determination by including the total time at the time of the input in the determination element.

According to the above configuration, since the judgment is performed by including the total time at the time of input in the judgment element, it becomes more difficult for the observer to specify the password.

The voice input device according to the nineteenth aspect of the present invention
In order to solve the above-mentioned problems, a speech recognition input means for recognizing a series of input speeches inputted for password input, converting the speech into speech data and outputting the speech data, and a specific word storage means for storing a plurality of specific words in advance And password identification means for detecting whether or not some or all of the plurality of specific words stored in the specific word storage means are included in the voice data output from the voice recognition input means. And a voice input time measuring means for measuring an input time interval of a plurality of specific words included in the voice data, wherein the password identifying means, together with the detection result, the input It is characterized in that the success or failure of the password input is determined by including the time interval in the determination element.

According to the above arrangement, even if the password is input by voice in a situation where the password can be heard by another person, the password is not specified by the other person and the user is not even aware that the password is input. Input can be realized.

That is, since whether or not a specific word is included in a series of input voices is detected and the input time interval is included in the determination element to determine whether or not the password input is successful, the observer is , You can't identify which word was the password, and you don't even notice that you're typing the password.

According to a twentieth aspect of the present invention, a voice input device
In order to solve the above problems, a voice input time measuring means for measuring a voice input time of voice input for password input, and a password for determining whether password input is successful or not, including the voice input time as a determination element And identification means.

According to the above configuration, even if the password is input by voice in a situation where the password can be heard by another person, the password is not specified by the other person and the user is not even aware that the password is being input. Input can be realized.

That is, since the voice input time of the input voice is measured and the voice input time is included in the determination element to determine the success or failure of the password input, the observer can determine which word was the password. I can't identify it, and I don't even know I'm entering my password.

The voice input device according to the twenty-first aspect of the present invention
In order to solve the above problem, in any one of claims 1 to 20, a password is identified using any one of a plurality of password identification methods,
The password identification method can be selectively changed.

According to the above configuration, since various determination methods can be switched and used, the password input method can be made more difficult for others to steal.

[0056]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to FIGS.

FIG. 1 is a block diagram showing a schematic configuration of the voice input device according to the present embodiment. The present voice input device can be applied to, for example, a private telephone terminal that has a voice password and can be connected to a public telephone. Exclusive connections are used by companies that have offices in various locations, because the cost of calls can be reduced even over long distances by using dedicated lines within the company.
In the use, for example, an outsider can operate a telephone from a dedicated telephone in the premises of the lobby.
To prevent this, the password input method of the voice input device can be applied.

This voice input device includes a microphone 11, voice recognition input means 1, specific word storage means 21, and password identification means 2.

The voice recognition input means 1 comprises software for recognizing voice and converting documents, and a computer. A word is cut out from a series of input voices input from the microphone 11, and a word is output as voice data for each word as a serial output port. And outputs it to the password identifying means 2. As this software, the above-mentioned conventional software can also be used.

The specific word storage means 21 stores one or more specific words to be a password in advance. The password identification unit 2 compares the content of the specific word storage unit 21 with the voice data input from the voice recognition input unit 1, and detects whether or not the voice data includes a specific word serving as a password. I do.

If the word in the voice data matches the specific word, a password determination signal indicating that the password has been detected is output. If the word does not match, a password determination signal indicating that the password has not been detected is output. Output. The password determination signal is connected to, for example, a telephone device (not shown), and the telephone device receives a password input, and is ready to receive a dial. The user who has input the regular password becomes usable. On the other hand, an outsider who does not know the password cannot input the password correctly, and cannot use the telephone.

The voice input device may be provided with a password input success / failure display means 50 for displaying the success or failure of the password input after a series of voice inputs is completed. FIG. 2A shows the password input success / failure display means 50.
As shown in (2), when the password input is completed normally, it is displayed that the password input is successful. This display may be changed from the current display to the next display as shown in (b) and (c) of FIG. 2 in addition to the display that is explicitly displayed as shown in FIG. FIG. 2B shows a current display before the password is input, and FIG. 2C shows a display for inputting a telephone number or the like as the next display after the password is input.

Next, a specific example of password input will be described. As an example, it is assumed that the word “telephone” is registered in the specific word storage unit 21. So he said, "Oh, you have to make a phone call."
If the password is entered, the word "telephone" matches, so that the password input has been completed normally, and the password input success / failure display means 50 indicates that the password input has been completed. This display may be a display as shown in FIG. 2A, as described above, or may be a display as shown in FIG.
The display may be changed to the display shown in (c).

At this time, the surrounding people merely think that the user has made a business call and do not think that the user has entered the password. Further, when the password needs to be input next time, the voice may be input in another sentence, for example, a sentence such as "I wonder if there is a public telephone somewhere" or "I forgot the mobile phone".

Further, the present voice input device may be provided with a password identification activation / termination means 3 for activating and / or terminating the operation of the password identification means 2.

The password identification activation / termination means 3 is provided with, for example, a switch 32 and can be operated as follows. That is, when the switch 32 is turned on, the password identification activation / termination means 3 generates an activation signal, whereby the password identification means 2 is put into the operating state.
When the switch 32 is turned off, the password identification activation /
The ending means 3 generates an ending signal, whereby the password identifying means 2 is put into a non-operation state.

According to the above configuration, it is possible to make it more difficult to determine what is a password. For example, the user turns off the switch 32 and says, "Oh, you have to make a phone call" to indicate to the observer that this is not acceptable as a password. next,
Turn on the switch 32 and say the words "Last month's phone bill took 30,000 yen", and by entering the password successfully, the previous "Ah, you have to call" is a password It can give the observer the impression that he / she did not, and give the material to deny that "phone" is a password.

A recording and reproducing means 4 for recording a series of input voices and reproducing the voices in the voice input device.
0, and the password identification means 2 outputs
It may be configured to perform password identification.

According to the above configuration, when the password is input, the sound input from the microphone 11 is transmitted to the recording / reproducing unit 40.
Once recorded. The recorded voice is reproduced after a predetermined time or according to a user's reproduction start instruction, and is input to the voice recognition input unit 1. by this,
Since the password input and the password identification can be shifted from each other, it is possible to make it more difficult to determine what the password is.

Further, according to the above configuration, the automatic gain adjustment can be performed over the entire input voice during the recording / reproducing process, so that the voice recognition input means 1 can receive an input at a stable volume. As a result, there is an effect that determination can be easily performed. The recording / reproducing means 40 can be constituted by a recording / reproducing IC element or the like.

An earphone 42 for listening to reproduced sound
And the configuration can be used in combination with the configuration including the switch 32. In this case, according to the playback sound heard by the earphone 42, the switch 32
By operating, the password identification unit 2 can be activated without the password being heard by the observer. Therefore, it becomes more difficult to decipher the password.

In the above configuration, the recording / reproducing means 40 is configured to record / reproduce the input voice. However, as shown by the broken line in FIG. The audio data may be recorded and reproduced. With this configuration, the same effect as described above can be obtained. Also, in this case, the user can output the synthesized voice by the voice synthesizer (not shown) from the earphone 42.

Further, in the above, the password identification activation /
The configuration in which the operation of the ending means 3 is controlled by the switch 32 has been described. However, as another configuration, a specific word storage means 31 for activating / ending the password identification activation / ending means 3 is provided. Can be.

The start / end specific word storage means 31 stores one or more start / end specific words serving as a password. Then, the voice data from the voice recognition input means 1 is compared with the activation / termination specific word, and when they match, the password identification activation / termination means 3 can be activated or terminated.

For example, if "Ah" is a specific word for activation and "Negative" is a specific word for termination in "Ah, it is necessary to make a phone call", a password is input by this input. Completes, ie succeeds,
Entering "Yes, I have to call" does not mean that the password has been entered. This can confuse the observer.

The specific word storage means 21 and / or
Alternatively, it is preferable that the contents of the start / end specific word storage unit 31 be initialized.

Initially set, the storage means 21.
If any particular word is stored in 31, the user can use the device from the point of acquisition by speaking the particular word. On the other hand, when the specific word is not initially set, in order for the user to use the device, some kind of password input means is separately provided, and the user needs to set and register the password using this. For this reason, in consideration of user's convenience, it is desirable that the specific words in the storage units 21 and 31 be initialized.

Further, the operation of the password identification activation / termination means 3 may be terminated by providing a timer (timer means) 33 or a counter (counting means) 34. The timer 33 measures the time after the password identification means 2 is activated by the password identification activation / termination means 3, and terminates the password identification operation after a predetermined time set in advance to a desired value. The counter 34
After the password identification unit 2 is activated by the password identification activation / termination unit 3, the number of input words is counted, and when a predetermined number of words set in advance to a desired value is input, the password identification operation is terminated.

According to the above-described configuration, it is possible to restrict the password input. Therefore, even if an observer attempts to enter an invalid password by mistake, even if the user accidentally hits a specific word for activation, the valid password is not entered. The password identification means 2 can be terminated before the input.

In the above example, since the word "telephone" is always used, the observer can easily specify the password. Therefore, as shown in the storage content A of FIG. 3, a plurality of specific words are stored in the specific word storage unit 21 and / or the start / end specific word storage unit 31. Thereby, the password identification means 2 or the password identification activation /
The ending means 3 can make a determination by comparing one of the voice data from the voice recognition input means 1 with a plurality of specific words. For example, if the storage content A is stored in the specific word storage means 21, a specific word is included in the input voice such as "Oh, I have to make a phone call" or "Oh, the clock is broken". Therefore, the password input is completed.

Further, a specific word discrimination rule storage means 22 for designating the input order of a plurality of passwords is provided, and the password identification means 2 determines whether or not the password is correctly input only when the input voice data has the specified order relationship. By adopting a configuration in which the password is identified as input, it is possible to make it more difficult to specify the password.

For example, the specific word discrimination rule storage means 22
It is assumed that the storage content B shown in FIG. The stored content B specifies the input order of inputting the word "cha" after the input of the word "telephone" or "clock", and the password identifying means 2 uses this order (rule).
Only when the password is input according to the above, it is determined that the password is correctly input.

Therefore, the input "ch" following "telephone""Ah, it is necessary to make a call"
In addition, the input "Ah, the clock has been broken" in which "Cha" is added after "Clock" is included in the specified order in the specified order, so that the rule of the stored content B is applied. When such an input is made, the password input is completed.

Of course, the above configuration may be used in combination with the configuration provided with the start / end specific word storage means 31. For example, if “Ah” is a specific word for activation among the activation / termination specific words, the password input is completed by inputting “Ah, it is necessary to make a phone call” or the like.

Further, the rules regarding the input order of a plurality of passwords as described above can be applied to a start / end specific word.

Further, when a plurality of specific words are used for a password as described above, depending on which of them is used,
The subsequent operation can be changed. That is, when the specific word “telephone” is recognized, the display shifts to the telephone display, and the telephone can be used. On the other hand, when the specific word “clock” is recognized, the clock can be displayed. In other words,
The password can also be used as the next operation command. However, in this case, it is necessary to keep in mind that the use of a word that means the subsequent operation content in the common sense for the password may cause the meaning of the password to be lost.

FIG. 6 is a flowchart showing the identification operation of the password identification means 2 in the configuration provided with the specific word discrimination rule storage means 22 described above. First, step S1
In, a specific word to be identified and stored in the specific word storage unit 21 is preset in advance. In this state, when a password is input and voice data is input to the password identification unit 2 (step S2), the specific word storage unit 21
Are extracted (step S3).

Then, the voice data input from the voice recognition input means 1 is compared with the target specific word in the specific word storage means 21 to determine which one of the specific words matches (step S4). If they do not match, the target specific word is updated (step S8), and the process returns to the stage before step S2.

On the other hand, if they match, the judgment rule of the specific word judgment rule storage means 22 is extracted (step S).
5) It is determined whether the password input has been completed in accordance with the rule (step S6). If the password input has been completed, the password determination signal is completed (step S7). If not completed, the target specific word is updated (step S8), and the process returns to the stage before step S2.

For example, the discriminating operation in the case where the storage contents of a specific word or the like are as shown in FIG. 3 will be described. 1/2 ”or“ Clock / 3/2 ”
Is the target specific word (step S1). When there is an input of "telephone" (step S2), "telephone" and "clock" are taken out (step S3), and since "telephone" matches (step S4), steps S5 to S5 are performed.
Proceed to 6. Here, from the specific word discriminating rule, it is found that the word is not completed yet but is a specific word number candidate continuing 2 (step S6), and the target specific word is the word “Cha / 2 / Complete”.
Is updated (step S8). When there is an input of "ch" (step S2), the password determination signal is completed through steps S3 to S7.

Here, the “identification target specific word” in step S1, the “target specific word” in step S3, and the “target specific word” in step S8 mean the same thing, such as a pointer indicating how far the determination has proceeded. It has become something.

The identification operation of the password identification means 2 in a configuration in which the specific word identification rule storage means 22 is not provided is as follows.
Except for the steps S5 and S6, the procedure is the same as that shown in the flowchart of FIG.

It is also possible to provide a specific word updating means 60 for setting and updating a specific word to be stored in the specific word storage means 21 and / or the start / end specific word storage means 31 in the present voice input device. ,preferable.

Setting of Specific Word by Specific Word Update Means 60
Updating can be realized by writing the voice data input via the voice recognition input means 1 into the specific word storage means 21. This setting / updating operation can be performed, for example, as follows. A setting / updating specific word storage unit (not shown) for storing a setting / updating specific word is provided. When the setting / updating specific word is “addition of specific word”, the setting / updating is performed. Speech input of "add specific word" is performed, and then a word of "fax" is input as a specific word to be added. Thereby, as shown in FIG. 4, the word "fax" can be added to the specific word storage unit 21.

The above setting / updating operation is desirably performed in a place where no one else is. In addition, if the specific word for setting / updating is a word such as “add specific word”, even if the existing password cannot be specified, the device may be stolen by the viewer adding a new password. You have to keep this in mind. Therefore, it is preferable that this specific word for setting / updating can also be changed.

To change a specific word in the case of a personally managed terminal, in addition to the above setting / updating operation, a part of the input speech including the specific word whose password has been successfully recognized is taken out according to the rules. A method of automatically changing the specific word that has succeeded can also be used. For example, when an input of “fax received” is input, the password identification unit 2 outputs a password determination signal upon detection of the specific word “fax”, and the specific word update unit 60 that receives the password determines the “receive”. Take out. As a result, "receiving" is overwritten in the specific word storage means 21 as a specific word,
As a result of the "fax" being deleted, the storage content A of the specific word storage means 21 is changed from the one shown in FIG. 4 to the one shown in FIG.

According to the above configuration, the next time a password is input, for example, "yearly gift received" may be input. Further, when the user inputs "the sea right below", the judgment result will differ depending on whether or not the audio data is created with a meaning. If there is no meaning, it will be recognized as a password because it has the same "was". If it has a meaning (including Kanji conversion),
Since "was" and "below" are different, they will not be recognized as passwords.

The above-mentioned password identification means 2, password identification activation / termination means 3, and specific word updating means 60
For example, as shown in FIG. 7, it can be realized by software processing by a microprocessor 4 incorporated in the present voice input device. Here, the password identification unit 2 is configured as shown in FIG.
When the password identification activation / termination means 3 is also used together with the activation / termination specific word storage means 31 or the like, the processing according to the processing shown in the flowchart of FIG. 6 is performed. . Further, the specific word updating means 60 performs the setting / update processing described above.

The above-mentioned specific word storage means 21, specific word discrimination rule storage means 22, and start / end specific word storage means 3
For example, as shown in FIG.
Can be realized by using a specific area of the RAM 5 connected to the. Timer 33 and counter 3
4 can be realized by a built-in clock and a built-in counter built in the apparatus, respectively, and the password input success / failure display means 50 can be realized by a display device such as an LCD.

In the conventional one-word password input, one word is input alone. However, when a series of inputs is performed as in the present invention, due to the nature of the continuous input, an arbitrary or arbitrary character before or after a specific word is input. By specifying the presence or absence of a specific word or the input interval of each word as a rule, the element can be reflected in the password input so that the observer does not know the element.

First, FIG. 8 shows an example in which a rule is established that a specific word follows an arbitrary word. In this example, the above rule is realized by providing a special specific word “?” That allows input of an arbitrary word in the specific word determination rule storage unit 22.

FIG. 9 shows an example in which an arbitrary word comes after a specific word as a rule, and FIG. 10 shows that a specific word comes before and after an arbitrary word. An example in the case of performing is shown. Also in these examples, by providing a special specific word “?”, Password input using an arbitrary word is realized.

Further, as shown in FIG. 11, by providing the voice input time measuring means 70 for measuring the non-voice period or the voice period, the time interval or the voice input time can be incorporated in the rule. The voice input time measuring means 70 can be realized by, for example, a built-in clock built in the device.

According to the above configuration, the success or failure of the password input is determined by the password identification means 2 depending on whether or not the measurement result of the voice input time measurement means 70 complies with the rule of the specific word discrimination rule storage means 22.

For example, FIG. 12 shows an example in which a rule that a specific word comes after a silence of one second or more is made.
In this rule, there is a one-second silence period before the specific word "telephone" in "yes, I have to make a phone call", and "Aye, I have to make a phone call (" ‥‥‥ ‥ ”represents a silent period)”, and is recognized as a normal password. Similarly, it is possible to make a rule that a silent state occurs after a specific word, and it is also possible to make a rule that a silent state occurs before and after a specific word.

On the contrary, as shown in FIG. 13, when the silent state continues for one second or longer, the initialization is performed, thereby
An input in which a specific word is uttered after silence for more than a second can be rejected as a password. In this example, if "Oh, I have to make a phone call" is uttered continuously, it is recognized as a correct input, while "Ah, I have to make a phone call" is accepted as a password Rejected.

Further, when a specific word is not specified and the voice input time is equal to or longer than a certain time, the voice input can be recognized as a password. in this case,
The password identification unit 2 makes a determination using only the time interval measured by the voice input time measurement unit 70 as a determination element. For example, FIG. 14 shows an example in which a sound period longer than 3 seconds is recognized as a correct password input.

As shown in FIG. 15, it is also possible to determine the success or failure of the password input by using the input time interval between each specific word of the voice input including a plurality of specific words as a determining factor.

Further, as shown in FIG. 16, the duration of a voice input including a specific word (with a non-voice period of 0.5 seconds or more as a delimiter) is measured, and the time becomes longer than or equal to a certain time. Only when the specific word can be recognized as a password. In this case, the duration of the voice is measured by the voice time input measuring means 70, and the measured duration is passed directly to the password identifying means 2, and the specific word discriminating rule storing means 22
Is determined according to the following.

As described above, by changing the contents of the specific word discriminating rule storage means 22, it is possible to switch and use various judging methods. Therefore, it is necessary to prepare a plurality of discriminating rules and switch between them. Thus, the password input method can be made more difficult for others to steal.

For example, in the contents of the specific word discriminating rule storage means 22 shown in FIG. 3, both the specific word numbers 1 and 3 have been recognized as a target, but input from only the number 1 is selectively effective. Then, only "I must call" is the correct password input, and "Clock" is not recognized.

Conversely, if only the specific word number 3 is valid, "clock" is recognized and "call must be" is not recognized as a success password.

In the present voice input device, the voice is input through the microphone 11, but any device capable of transmitting and inputting voice can be used, such as a wireless transmission such as a telephone line, a mobile phone, or an MCA, or a CATV line. Anything is fine.

Further, there are a wide variety of devices to which the present device can be applied, such as mobile phones, portable information terminals, security locks, and commerce devices, where the device itself may be stolen.

[0115]

According to the first aspect of the present invention, there is provided a voice input device for recognizing a series of input voices input for inputting a password, converting the input voices into voice data, and outputting the voice data. A specific word storage unit that stores one or more specific words serving as a password in advance, and a specific word stored in the specific word storage unit are included in voice data output from the voice recognition input unit. Password identification means for detecting whether a password is included in a series of input voices.

Therefore, even if the password is input by voice in a situation where the password can be heard by another person, the input can be realized without the password being specified by the other person and without being aware that the password is being input. This has the effect.

According to a second aspect of the present invention, as described above, the voice input device of the first aspect is provided with a password identification activation / termination means for activating and / or terminating the operation of the password identification means. is there.

Therefore, it is possible to make it more difficult for the observer to determine what the password is.

According to the third aspect of the present invention, as described above, in the configuration of the first or second aspect, while recording a series of input voices or voice data and recording / reproducing the voice or voice data. Means, and the password identification means is configured to identify the password by the reproduced voice or voice data.

[0120] Therefore, the time of inputting the password and the time of identifying the password can be shifted, so that it is more difficult to determine what the password is.

According to a fourth aspect of the present invention, as described above, in the configuration of the third aspect, during playback, the password identification means is activated and / or terminated by the password identification activation / termination means. Configuration.

Therefore, the decryption of the password can be made more and more difficult for the observer.

According to a fifth aspect of the present invention, as described above, in the configuration of the third or fourth aspect, the time from the time of recording to the start of reproduction is fixed in advance for the input or recorded voice or voice data. Or any configuration that can be set arbitrarily.

Therefore, the operation can be simplified if the time is fixed in advance. Further, by changing the time, it is possible to make it more difficult for the observer to decipher the password.

According to a sixth aspect of the present invention, there is provided a voice input device according to any one of the first to fifth aspects,
A plurality of specific words are stored in the specific word storage unit, and the password identification unit includes part or all of the plurality of specific words in the voice data output from the voice recognition input unit. Is detected, the password is correctly input.

Therefore, by having a plurality of password target words and entering a password by matching some or all of them, it becomes difficult to specify a particular word no matter how many times it is observed. be able to.

According to a seventh aspect of the present invention, as described above, in the configuration of the sixth aspect, according to the type of the specific word included in the voice data detected by the password identification means, The contents that can be accessed after entering the password are different.

Therefore, the subsequent operation can be changed depending on which specific word is used.

According to the eighth aspect of the present invention, the voice input device according to the sixth aspect further includes a specific word discrimination rule storage unit for designating an input order of a plurality of passwords.
The password identification unit is configured to identify that the password has been correctly input only when the input voice data has the specified order relationship.

Therefore, by specifying that the password input is completed only when there is a relationship specified in the input order of a plurality of passwords, it is possible to further specify the password.

According to a ninth aspect of the present invention, as described above, in the configuration of the second aspect, the activation / termination means for activating and / or ending the password identification activation / termination means in accordance with the voice input. This is a configuration including an end specific word storage unit.

Therefore, if the password identification means adopts the input method activated by voice input, even if the password is uttered before activation, it is not accepted, and it is possible to confuse the observer.

According to a tenth aspect of the present invention, as described above, in the configuration of the second aspect, after the password identification means is activated by the password identification activation / termination means, time is measured and a predetermined time elapses. This is a configuration provided with a timer for terminating the password identification operation later.

Therefore, a password input method having an allowable time for password input can be adopted, and it is possible to prevent an observer from repeatedly inputting the password.

According to the eleventh aspect of the present invention, as described above, in the configuration of the second aspect, after the password identification means is activated by the password identification activation / termination means, the number of input words is counted. And a counting means for terminating the password identification operation when a predetermined number of words are input.

Therefore, a password input method having an allowable number of words can be used for password input, and it is possible to prevent an observer from repeatedly inputting the password.

According to a twelfth aspect of the present invention, as described above, the specific word storage means and / or activation / termination by the input or reproduced audio data in any one of the first to eleventh aspects. This is a configuration including specific word updating means for changing the contents of the specific word storage means.

Therefore, the specific word to be stored can be changed, so that it becomes more difficult to specify the password.

According to a thirteenth aspect of the present invention, as described above, in the configuration of any one of the first to eleventh aspects, at the time of inputting the password, the input is performed immediately after the specific word used as the password or after a few words. The configuration is provided with specific word updating means for changing the contents of the specific word storage means according to words.

Therefore, the word pronounced after the password at the time of inputting the password becomes the next password, and the password is changed every time, so that it becomes more difficult to specify the password.

According to a fourteenth aspect of the present invention, as described above, the success or failure of the password input is displayed after a series of voice inputs is completed. This is a configuration including display means.

Therefore, the operator can know the success or failure of the password input by the display.

According to a fifteenth aspect of the present invention, there is provided the voice input device according to any one of the first to fourteenth aspects, wherein the contents of the specific word storage means and / or the start / end specific word storage means are as follows: This is the initial configuration.

Therefore, the user can use the apparatus from the time of acquisition by inputting the specific words that are initially set by voice, thereby improving the user's convenience.

As described above, the voice input device of the present invention recognizes a series of input voices input for inputting a password, converts the input voice into voice data, and outputs the voice data. A specific word storage unit that stores in advance one or more specific words, and whether the specific word stored in the specific word storage unit is included in the voice data output from the voice recognition input unit. And a password identification means for detecting whether or not there is a voice input only before or after the specific word or before and after the specific word. It is a configuration provided.

Therefore, even if the password is input by voice in a situation where the password can be heard by another person, the input can be realized without the password being specified by the other person and without being noticed that the password is being input. This has the effect.

The speech input device according to the seventeenth aspect of the present invention, as described above, is characterized in that, in the configuration of the sixteenth aspect, the time interval between the input of a specific word and a word input before and / or after the specific word is input. And a password identification unit configured to perform the determination by including the time interval in a determination element.

Therefore, since the determination is performed by including the time interval in the determination element, it is possible to make it more difficult for the observer to specify the password.

As described above, in the speech input device according to the eighteenth aspect of the present invention, in the configuration of the sixteenth aspect, the total time of input of the specific word and the voice input before and / or after the specific word is input. The password identification means is configured to make a determination by including the total time at the time of the input in a determination element.

Therefore, since the determination is performed by including the total time at the time of input in the determination element, it is possible to make it more difficult for the observer to specify the password.

The voice input device according to the nineteenth aspect of the present invention comprises a voice recognition input means for recognizing a series of input voices input for inputting a password, converting the input voice into voice data, and outputting the voice data. A specific word storage unit that stores a specific word in advance, and a part or all of the plurality of specific words stored in the specific word storage unit are included in voice data output from the voice recognition input unit. And a voice input time measuring means for measuring input time intervals of a plurality of specific words included in the voice data. The means is configured to include the input time interval in the determination element together with the detection result, and determine the success or failure of the password input.

Therefore, even if the password is input by voice in a situation where the password can be heard by another person, the input can be realized without the password being specified by the other person and without even being aware that the password is being input. This has the effect.

According to a twentieth aspect of the present invention, as described above, the voice input time measuring means for measuring the voice input time of the voice input for inputting the password, and the voice input time are included in the determination element. And a password identification means for determining whether or not the password input is successful.

Therefore, even if the password is input by voice in a situation where the password can be heard by another person, the input can be realized without the password being specified by the other person or even without being aware that the password is being input. This has the effect.

As described above, in the voice input device according to the twenty-first aspect of the present invention, in any one of the first to twentieth aspects, the identification of a password is performed by using any one of a plurality of password identification methods. The password identification method can be selectively changed.

Therefore, since various judgment methods can be switched and used, the password input method can be made more difficult for others to steal.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a schematic configuration of a voice input device according to an embodiment of the present invention.

FIGS. 2A to 2C are diagrams showing display examples of a password input success / failure display means provided in the voice input device.

FIG. 3 is a diagram exemplifying storage contents of a specific word storage unit and a specific word discrimination rule storage unit provided in the voice input device.

FIG. 4 is a diagram showing an example of updating the storage contents of the specific word storage means and the specific word discrimination rule storage means.

FIG. 5 is a diagram showing another example of updating the storage contents of the specific word storage means and the specific word discrimination rule storage means.

FIG. 6 is a flowchart illustrating an identification operation of a password identification unit provided in the voice input device.

FIG. 7 is a block diagram illustrating a configuration example of the voice input device.

FIG. 8 is a diagram showing another storage content of the specific word discrimination rule storage means.

FIG. 9 is a diagram showing still another storage content of the specific word discrimination rule storage means.

FIG. 10 is a diagram showing still another storage content of the specific word discrimination rule storage means.

FIG. 11 is a block diagram showing a configuration in which a voice input time measuring means is provided in the voice input device.

FIG. 12 is a diagram showing still another storage content of the specific word discrimination rule storage means.

FIG. 13 is a diagram showing still another storage content of the specific word discrimination rule storage means.

FIG. 14 is a diagram showing still another storage content of the specific word discrimination rule storage means.

FIG. 15 is a diagram showing still another storage content of the specific word discrimination rule storage means.

FIG. 16 is a diagram showing still another storage content of the specific word discrimination rule storage means.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Voice recognition input means 2 Password identification means 3 Password identification start / end means 21 Specific word storage means 22 Specific word discrimination rule storage means 31 Start / end specific word storage means 33 Timer (time measuring means) 34 Counter (Counting means) 40 Recording Reproduction means 50 Password input success / failure display means 60 Specific word update means

Continued on the front page (51) Int.Cl. ⁶ Identification symbol FI H04M 1/66 H04M 1/66 B // H04L 9/32 H04L 9/00 673D

Claims (21)

[Claims] 1. A voice recognition input means for recognizing a series of input voices input for inputting a password, converting the input voice into voice data, and outputting the voice data, and a specific word for storing one or more specific words serving as a password in advance. Storage means, in the voice data output from the voice recognition input means, detects whether the specific word stored in the specific word storage means is included, thereby, a series of input speech A voice input device comprising: a password identification unit for identifying whether or not a password is included therein. 2. The voice input device according to claim 1, further comprising password identification activation / termination means for activating and / or ending the operation of said password identification means. 3. A recording / reproducing means for recording a series of input voices or voice data and reproducing said voice or voice data, wherein said password identification means identifies a password based on said reproduced voice or voice data. The voice input device according to claim 1 or 2, wherein: 4. The voice input device according to claim 3, wherein said password identification means is activated and / or terminated by said password identification activation / termination means during reproduction. 5. The voice according to claim 3, wherein the time from the time of recording to the start of reproduction of the input or recorded voice or voice data is fixed in advance or can be set arbitrarily. Input device. 6. A plurality of specific words are stored in said specific word storage means, and said password identification means includes a part or all of said plurality of specific words in voice data output from said voice recognition input means. By detecting that is included,
The voice input device according to any one of claims 1 to 5, wherein the voice input device identifies that the password has been correctly input. 7. A method according to a type of a specific word included in voice data, detected by said password identification means.
7. The voice input device according to claim 6, wherein accessible contents are different after inputting the password. 8. A specific word discriminating rule storing means for designating an input order of a plurality of passwords, wherein said password identifying means correctly inputs a password only when input voice data has the specified order relationship. The voice input device according to claim 6, wherein the voice input device is identified as having been performed. 9. The voice input device according to claim 2, further comprising a start / end specific word storage unit for starting and / or ending the password identification start / end unit in accordance with a voice input. 10. After the password identification means has been activated by the password identification activation / termination means, time measurement is performed.
3. The voice input device according to claim 2, further comprising a timer for terminating a password identification operation after a predetermined time has elapsed. 11. After the password identification means is activated by the password identification activation / termination means, the number of input words is counted, and when a predetermined number of words is input, the counting means for terminating the password identification operation is provided. The voice input device according to claim 2, further comprising: 12. A system according to claim 1, further comprising a specific word updating means for changing the contents of said specific word storage means and / or the start / end specific word storage means in accordance with input or reproduced voice data. The voice input device according to claim 1. 13. When inputting a password, a word input immediately after or a few words after a specific word serving as a password is used.
The voice input device according to claim 1, further comprising a specific word updating unit that changes contents of the specific word storage unit. 14. The voice input device according to claim 1, further comprising a password input success / failure display means for displaying the success or failure of the password input after a series of voice inputs is completed. 15. The voice input device according to claim 1, wherein the contents of the specific word storage means and / or the start / end specific word storage means are initialized. . 16. A voice recognition input means for recognizing a series of input voices input for inputting a password, converting the input voice into voice data, and outputting the voice data, and a specific word for storing one or more specific words serving as a password in advance. Storage means, and detects whether or not the specific word stored in the specific word storage means is included in the voice data output from the voice recognition input means, and only before the specific word. Or a password identification unit that detects whether there is a voice input only at the back, or before or after, and thereby determines whether or not the password input is successful. 17. A voice input time measuring means for measuring a time interval at the time of inputting a specific word and a word input before and / or after the specific word, wherein the password identifying means determines the time interval. 17. The voice input device according to claim 16, wherein the determination is performed by including the determination element. 18. A voice input time measuring means for measuring a total time of inputting a specific word and a voice input before and / or after the specific word, wherein the password identifying means comprises: 17. The voice input device according to claim 16, wherein the determination is performed by including the total time in the determination element. 19. A voice recognition input means for recognizing a series of input voices input for inputting a password, converting the input voice data into voice data, and outputting the voice data; a specific word storage means for storing a plurality of specific words in advance; Password identification means for detecting whether or not some or all of the plurality of specific words stored in the specific word storage means are included in the voice data output from the recognition input means; A voice input time measuring unit for measuring an input time interval of a plurality of specific words included in the voice input device, wherein the password identification unit determines the input time interval together with the detection result. A voice input device for determining whether password input is successful or not, including in an element. 20. A voice input time measuring means for measuring a voice input time of a voice input for inputting a password, and a password identifying means for determining whether password input is successful or not by including the voice input time in a determination element. A voice input device comprising: 21. The method according to claim 1, wherein the password is identified by using one of a plurality of password identification methods, and the password identification method can be selectively changed. The voice input device according to claim 1.