JP3728921B2 - Voice command terminal device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a voice command terminal device such as a voice recognition card.
[0002]
[Prior art]
In recent years, voice recognition technology has been put into practical use. There are two types of speech recognition technologies: those targeting unspecified speakers and those targeting specific speakers. For example, it is known that when an unspecified speaker pronounces a name in front of a public facility having a voice recognition function, the name is recognized and a corresponding process is executed.
[0003]
On the other hand, for a specific speaker, a specific person's voice is registered, and voice recognition is performed only on the registered person, and processing corresponding to the recognized voice is executed. It is.
[0004]
[Problems to be solved by the invention]
However, the conventional speech recognition apparatus for unspecified speakers described above targets a large number of unspecified users, so the number of recognizable words is limited or the recognition rate is good when trying to increase the number of words It has not reached the practical range. In addition, the above-described speech recognition apparatus for a specific speaker has a problem that it can be used only by a limited number of specific persons with a high recognition rate.
[0005]
Furthermore, since the place where voice recognition is performed is fixed, it is not suitable for the case where it is necessary to maintain confidentiality because it is affected by surrounding noise, etc., or it must be pronounced in an environment where other people may exist. There was a problem.
[0006]
An object of the present invention is to provide a voice recognition function by a specific speaker with a high recognition rate to a terminal possessed by each user so that the user can perform voice input at a desired location, while maintaining confidentiality. It is to be able to recognize an unspecified number of voices with a very high recognition rate.
[0007]
[Means for Solving the Problems]
The voice command terminal device according to the present invention includes a voice recognition unit for recognizing an input voice and registration for registering command information corresponding to the voice recognized by the voice recognition unit for each device type in a form different from the voice. And the command information of the other device registered in the registration unit when the information indicating the approach to the other device that receives the command information and executes the corresponding process is acquired from the other device. Transmission means for transmitting to the apparatus.
[0010]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 is a diagram showing a configuration of a voice recognition card to which a voice command terminal device of the present invention is applied. Reference numeral 10 in the figure is a CPU as a control unit that controls the entire voice recognition card. Reference numeral 11 denotes a microphone as a voice input unit, and reference numeral 12 denotes a keyboard, which includes keys for inputting various commands including voice command keys and transmission keys. Reference numeral 13 denotes a communication interface as transmission means when transmitting command information corresponding to voice input by the user to another device such as a vending machine. Reference numeral 14 denotes a plurality of types of target models that can be selected by the user. A display unit 14 for displaying a menu, 15 is a speech recognition unit (speech recognition means) for recognizing speech input by a user, 16 is a flash ROM used as a speaker registration dictionary, and 17 is It is a command table used for obtaining command data from a sound input by a search by the CPU 10.
[0011]
FIG. 2 is a diagram showing the configuration of the command table shown in FIG. 1, and a plurality of model codes selectable by the user, command data (command information) used in each model, and voice data corresponding to each command data Are stored in association with each other. In FIG. 2, for example, three command data and voice data are shown as an example for the model code of the vending machine A for ticket sales.
[0012]
FIG. 3 is a diagram illustrating an example of a menu displayed on the display unit 14. Vending machine A, 2. Vending machine B, 3. Door A, 4. Door B, 5. TV is shown as an example.
[0013]
Details of the voice registration process for the voice recognition card will be described below with reference to the flowchart shown in FIG.
First, when a voice command key is pressed, candidates for the target model as shown in FIG. 3 are displayed on the display unit 14 (step S1). The user looks at this menu and selects, for example, a vending machine A for ticket purchase (step S2), and then inputs a destination such as “Tokyo” or “Osaka” via the microphone 11 by voice (step S2). Step S3). Then, the speech recognition unit 15 recognizes the input speech (step S4). Next, command data corresponding to the voice data recognized by scanning the command table 17 is retrieved (step S5), and the retrieved corresponding command data is stored in the selected model (here, vending machine A). Along with the model code, it is stored in the flash ROM 16 in the form of text data (step S6). Next, it is determined whether there is other voice data to be registered (step S7). If there is another voice data, the process returns to step S3 to input voice. For example, the user can input a plurality of voice data corresponding to a command to be executed in advance for each target model (for example, for one day). As a result, the command data corresponding to the voice data can be selected according to the selected model.
[0014]
Also, the voice data and the voice data converted from the voice data (for example, “PC” is converted to “personal computer”) are associated with each other and registered in the flash ROM 16, and the voice “PC” is recorded. “Personal computer” may be registered when input.
[0015]
FIG. 5 is a flowchart showing details of processing when command data registered by the above method is transmitted. First, as shown in FIG. 7, the user 99 receives model approach information from the vending machine 101 by bringing the voice recognition card 100 close to the vending machine 101 for ticket purchase, for example, within 3 m, or the user 99 99 determines whether or not the transmission key has been pressed (steps S8 and S9), and if either determination is YES, the voice recognition card 100 includes command data corresponding to the voice data registered in advance. The model code is transmitted to the vending machine by non-contact means such as infrared rays (step S10). When the command data is normally recognized by the vending machine 101, a response transmission is returned to the voice recognition card 99 side, so step S11 is YES, and it is determined whether there is other command data to be transmitted. (Step S12). If NO, the process ends. If YES, the process returns to step S8 to determine whether there is model approach information (step S8) or whether the transmission key is pressed (step S9). On the other hand, if the command data is not recognized normally, no response is transmitted and an error is displayed (step S13).
[0016]
Next, details of command reception processing by the vending machine 101 will be described with reference to the flowchart of FIG. The vending machine 101 always transmits model approach information (step S20). When this is recognized by the voice recognition card 100, the command data and the model code are transmitted from the voice recognition card 100 side. It is determined whether or not command data whose model code matches its model code has been received (step S21). When command data that matches the model code is received, the command data is analyzed (step S22), and it is determined whether or not an article (ticket) corresponding to the command data can be provided to the user 99 (step S23). ). Here, if the user has inserted a predetermined amount of money and the destination is available for sale, the answer is YES, and then a response is sent to the voice recognition card 100 (step S24), followed by the article. After performing the corresponding process of providing (ticket) to the user 99 (step S25), the process returns to step S20. If NO in step S23, an error display (step S26) is performed and the process returns to step S20.
[0017]
According to the above-described embodiment, the user uses the portability of the voice recognition card to perform voice registration in advance at a desired place where no other person is present, and then performs voice recognition at the place where the vending machine is installed. Command data can be transmitted by carrying the card. Since the speech recognition method in this case is intended for a specific speaker only for a user who possesses a speech recognition card, an unspecified number of speeches can be recognized with a very high recognition rate while maintaining secrecy. As a result, in the case of ticket vending machines, people who have low vision or those who are unfamiliar with machine scanning need help without having to look at the price list and without having to operate the keys. You can easily and reliably purchase tickets without having to pay for them.
[0018]
As a command transmission method, light or electromagnetic coupling may be used as non-contact means in addition to infrared rays. Alternatively, a dedicated card reader may be provided in the vending machine and an electrical contact method may be used. Moreover, it is not limited to a vending machine for purchasing tickets, but may be any other vending machine for purchasing beverages.
[0019]
A modification in which the above-described voice recognition card is applied to an automatic door lock will be described below with reference to FIG.
First, the user 199 registers a predetermined password in advance in the voice recognition card 200 at a desired location. For example, if “Suzuki Ichiro” is entered, it is converted into a password “54184973” and registered. The voice registration method can be performed in accordance with the method described above.
[0020]
Next, when the user 199 enters / enters through the door lock, the user 199 speaks to the voice recognition card 200 as “Suzuki Ichiro”.
Next, the voice recognition card 200 is brought close to the auto-lock device 201. If it falls within the reception range, the password “54184973” is input to the auto-lock device 201.
[0021]
Next, the auto-lock device 201 recognizes this code number, so that the door lock is released and entry / entry from the door 202 becomes possible.
According to the above-described modification, an unspecified number of voices can be recognized with a very high recognition rate while maintaining confidentiality. In addition, since it is not necessary to memorize a long password, entry / entry operation is simplified. In addition, key operations such as an auto-lock device are not required, and the convenience is greatly improved, and it can also be used by persons with disabilities and the elderly. In addition, because it is a specific speaker recognition, it is impossible for unauthorized persons to obtain a voice recognition card and speak "Suzuki Ichiro" so that unauthorized entry / entry is not possible. Greatly improved.
[0022]
As a further modification, by providing the voice recognition card with a dictionary function, it can be used for a remote control of a home appliance, a telephone directory, or the like. For example, “Wow Wow” is input by voice using a TV remote control or the like, and this is converted to “BS5” and transmitted to the TV. It is also possible to transmit it to the telephone after converting it to “0449353111”.
[0023]
【The invention's effect】
According to the present invention, each terminal device possesses a voice recognition function by a specific speaker having a high recognition rate, and by approaching this terminal device to another device, it corresponds to the other device approaching. The registered command information can be transmitted to the other device, and the command information can be accurately transmitted corresponding to the other device to execute processing by the other device.
[Brief description of the drawings]
FIG. 1 is a diagram showing a configuration of a voice recognition card to which a voice command terminal device of the present invention is applied.
FIG. 2 is a diagram showing a configuration of a command table of a voice recognition card.
FIG. 3 is a diagram showing an example of a menu displayed on the display unit of the voice recognition card.
FIG. 4 is a flowchart for explaining details of a voice registration process for a voice recognition card.
FIG. 5 is a flowchart showing details of processing when transmitting registered command data.
FIG. 6 is a flowchart for explaining details of command reception processing by the vending machine;
FIG. 7 is a diagram for explaining the operation when a voice recognition card is applied to a vending machine.
FIG. 8 is a diagram for explaining the operation when a voice recognition card is applied to an auto-lock device.
[Explanation of symbols]
10 ... CPU,
11 ... Mike,
12 ... Keyboard,
13. Communication interface,
14 ... display part,
15 ... voice recognition unit,
16 ... Flash ROM,
17 ... Command table.

Claims (2)

Speech recognition means for recognizing the input speech;
Registration means for registering the command information corresponding to the voice recognized by the voice recognition means for each device type in a form different from the voice;
Transmits information indicating access to another device that executes processing corresponding to receive the instruction information when acquired from the other device, the command information of the other devices that are registered in the registration unit to the other device Sending means to
A voice command terminal device comprising:   2. The command information corresponding to the recognized voice is selected according to another device selected from the plurality of other devices. Voice command terminal device.

Images