JP2003195885A - Communication device and its control method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accelerate processing by lightening the load of speech recognition processing and to improve speech recognition precision. <P>SOLUTION: A communication device (100) which is equipped with a speech recognition means (102) for recognition an inputted speech and can be controlled according to the output of the speech recognition means (102) comprises a storage means (104) for storing a plurality of kinds of dictionaries for speech recognition and a selecting means (103) for selecting a dictionary for speech recognition corresponding to the operation mode of the communication device (100) from the storage means (104) and supplies the selected dictionary to the speech recognition means (102). <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention relates to a communication device having a voice recognition function.

[0002]

2. Description of the Related Art In a communication device such as a telephone, a facsimile device, or a complex device of these devices, there is a device for recognizing a user's voice for operating the device or inputting a connection destination, instead of inputting keys on an operation section. Proposed.

Speech recognition processing is generally performed as follows.
First, the user's voice input from the microphone,
It is converted into a digital signal by an A / D converter and subjected to a predetermined filtering process for the purpose of removing noise components. Further, the frequency spectrum of the input voice data is calculated, and pattern matching with a standard voice pattern prepared in advance is performed. Then, in this pattern matching, the most similar standard voice among the standard voice patterns is output as the recognition result.

In recent years, in addition to the telephone function and the facsimile communication function, the function of transmitting and receiving an electronic mail (E-mail) and the Web by connecting to the Internet through a telephone line are provided.
A so-called multi-function communication device having a page browsing function (browser function) has been proposed.

As the number of functions increases, the operation may become complicated. That is, since the keys of the operation unit of a communication device are generally limited as compared with a keyboard of a personal computer or the like, complicated key operation is required when inputting an email address or an internet address.

Under these circumstances, a communication device having a voice recognition function has recently been proposed. According to the voice recognition function, it becomes possible to specify a connection destination by voice, and the operability is expected to be improved.

[0007]

However, since the conventional voice recognition function performs pattern matching between all the standard voice patterns registered in the voice recognition dictionary (voice dictionary) and the input voice pattern, the voice dictionary is used. There is a problem that the processing load increases as the number of standard voice patterns registered in is increased.

Further, since similar voices are searched for in the pattern matching, when the number of standard voice patterns increases, the possibility of selecting an incorrect standard voice pattern increases.
There is a problem that the accuracy of recognition decreases.

Further, when a destination has both a facsimile number and a telephone number, if the telephone number and the facsimile number of the destination are registered with the same destination abbreviation, the destination is simply There is also a problem that it is not possible to determine which one of the telephone and the facsimile is to be called by simply inputting the abbreviation. Therefore, in any case, it is necessary to select either telephone or facsimile by voice input or key input.
If an e-mail address is further registered in the recipient's abbreviation, it is necessary to select whether the function to be used is a telephone or facsimile and also an e-mail transmission / reception function.

The present invention has been made to solve the above problems, and in a communication device having a voice recognition means, the load of voice recognition processing is reduced, the processing speed is increased, and the voice recognition accuracy is improved. The purpose is to improve.

Further, according to the present invention, in a multi-function (including a call function, a facsimile communication function, an e-mail transmission / reception function, etc.) communication apparatus, it is possible to determine which function is desired to be used only by voice inputting the other party's abbreviation. It is possible to simplify the operation procedure.

[0012]

According to the present invention, there is provided a voice recognition device for recognizing an input voice, which is a communication device which can be controlled according to the output of the voice recognition device. It is characterized by further comprising: storage means for storing a recognition dictionary; and selection means for selecting a voice recognition dictionary according to the operation mode of the communication device from the storage means and supplying it to the voice recognition means. A communication device is provided.

Further, according to the present invention, there is provided a method of controlling a communication device, comprising a storage means for storing a plurality of types of speech recognition dictionaries, the determination step of determining an operation mode of the communication device, and the storage means. From the selection step of selecting a voice recognition dictionary according to the operation mode determined in the determination step, supplying to the voice recognition means, a voice input step of inputting voice, and the input voice, A voice recognition step of performing a voice recognition process using the voice recognition dictionary supplied in the selection step; and a step of executing a process in the operation mode based on an output of the voice recognition process. A method of controlling a communication device is provided.

Further, according to the present invention, a call means for making a call via a telephone line, a facsimile communication means for making a facsimile communication via a telephone line, and a data communication means for making a data communication via the Internet, A communication device comprising: a storage unit that describes a correspondence relationship between a connection destination abbreviation and information that identifies the connection destination; a voice input unit that inputs a voice; and a voice recognition unit that recognizes the input voice. And a determination means for determining which one of the communication means is to be used, the output of the voice recognition means and each connection destination abbreviation in the storage means are compared, and the connection destination is determined according to the comparison result. Collation means for selecting an abbreviation,
There is also provided a communication device, characterized in that the collation means includes control means for limiting collation candidates of each connection destination abbreviation in the storage means according to the determination result of the determination means.

Further, according to the present invention, storage means describing the correspondence between the abbreviation of the connection destination and the information for specifying the connection destination, the communication means for making a call through the telephone line, and the communication line for the telephone line are used. A method for controlling a communication device, comprising: a facsimile communication means for performing facsimile communication by means of facsimile communication; and a data communication means for performing data communication via the Internet, comprising a voice input step of inputting voice, and recognizing the input voice. Voice recognition step, a determination step of determining which one of the above-mentioned communication means is used, the recognition result obtained in the voice recognition step and each connection destination abbreviation in the storage means are compared. And a collating step of selecting a connection destination abbreviation according to the collating result, the collating step according to the determination result obtained in the determining step. The communication control method characterized by comprising the control step of limiting the matching candidates for each destination referred in the storage means is also provided.

[0016]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings.

(Embodiment 1) FIG. 1 is a block diagram of a communication device in this embodiment. Each block in this figure will be described below.

The communication device 100 is connected to another telephone or a facsimile device via the public line 113 and is capable of making a call or transmitting / receiving facsimile data. Further, it is connected to an Internet service provider (hereinafter referred to as ISP) via the public line 113, and it is possible to send and receive e-mail and browse Web pages via the ISP. That is, the communication device 100 has a call function for making a call via a telephone line, a facsimile communication function for performing facsimile transmission / reception, and a data communication function for performing data communication (including email transmission / reception and Web browsing) via the Internet. Have

The control unit 114 is composed of a CPU, a ROM storing a program, etc. (not shown), and controls each block in accordance with the program stored in the ROM.
Furthermore, it controls the entire communication device.

Reference numeral 101 is a microphone for inputting a user's voice and giving the input voice as an analog signal to the voice recognition unit 102.

The voice recognition unit 102 is an A / D converter,
It is composed of a signal processing circuit such as a DSP (both not shown), recognizes the user's voice input from the microphone 101, and outputs the recognition result to the control unit 114.

The user's voice input from the microphone 101 is converted into a digital signal by an A / D converter, and a predetermined filtering process for removing a noise component is performed in the signal processing circuit. Then, the frequency spectrum of the audio signal input in the signal processing circuit is calculated to generate an input audio pattern. further,
Pattern matching between this input voice pattern and the standard voice pattern given from the dictionary selection unit 103 is performed, and the most similar standard voice from the voice recognition dictionary (voice dictionary) is output as a recognition result.

Under the control of the control unit 114, the dictionary selection unit 103 selects a predetermined voice dictionary from a plurality of voice dictionaries registered in the voice dictionary memory 104, and outputs a standard voice pattern in the voice dictionary as a voice. This is given to the recognition unit 102.

The voice dictionary memory 104 includes a voice recognition unit 10.
This is a memory in which a standard voice pattern to be recognized in the second pattern matching is stored as digital format data. The structure of the voice dictionary stored in the voice dictionary memory 104 will be described later.

The operation unit 105 includes a dial key for inputting a connection destination address, a one-touch key for calling a pre-registered connection destination address, a start key for activating a transmission / reception operation, a stop key for stopping the operation, and a communication key of this communication device. It is composed of a mode setting key for setting the operation mode, and outputs information on the pressed state to the control unit 114. The user may operate the communication device 100 by using the operation unit 105, or may input the voice from the microphone 101 and operate the communication device 100 by voice recognition.

The display unit 106 is composed of an LCD (liquid crystal display), an LED (light emitting diode), etc., and the control unit 1
The operation state and the like of the communication device 100 are displayed under the control of 14. Further, it is possible to display the web page information loaded via the Internet and the content of the received e-mail. Further, it is also used to display the type of the voice dictionary selected by the dictionary selection unit 103 and inform the user of the recognizable voice type.

The reading unit 107 is a CCD (Charge Coupled).
Device), CS (contact type image sensor), or other photoelectric conversion means, which reads a document under the control of the control unit 114 and sequentially stores the image data as digital data in the data memory 109. Also,
It has an A / D conversion circuit, a shading correction circuit, a resolution conversion circuit, an edge enhancement circuit, a γ correction circuit, and the like, and performs binary or multivalued image processing on the read image data.
Further, it has a stepping motor for driving a document feeding / discharging roller, a driver circuit for driving the motor, a sensor circuit for detecting the document width and the presence or absence of the document, and the like based on the control of the control unit 114. Carry out.

The recording unit 108 includes a printer head such as a BJ (Bubble Jet (registered trademark)), a stepping motor that drives a paper feed / discharge roller, a driver circuit that drives these, and a sensor circuit that detects the recording paper width and the presence or absence of the recording paper. Under the control of the control unit 114, the facsimile received image, the Web page, etc. are printed out on the recording paper.

The data memory 109 is composed of a memory such as a DRAM and stores the image data read by the reading unit 107 and the facsimile data received from the connected facsimile via the public line 113. The data memory 109 also stores e-mail information created during e-mail transmission, web page information received from the ISP of the connection destination, and the like.

The address book 110 is composed of a nonvolatile memory such as SRAM or flash memory backed up by a battery or the like, and has a telephone (FAX) number or E of a connection destination.
Email address, internet address (eg UR
The connection destination abbreviation is added to L) and stored. When searching the address book, the connection destination address is searched based on the connection destination abbreviation. Furthermore, device-specific information, ISP phone number,
It also stores communication history.

The call unit 111 is a so-called handset, and is composed of a microphone and a speaker for making a call.

The communication control unit 112 is composed of a modem that modulates and demodulates communication data, a line supplemental circuit, a Ci (call signal) detection circuit, a dial number transmission circuit, etc., and under the control of the control unit 114, the communication device 100. And the telephone / facsimile device / ISP that is the connection destination to the public line 113
To connect and receive communication data.

A call is made by connecting to the call section 111 and the public line 113. In facsimile communication, communication data is transmitted and received according to the G3 facsimile procedure after connecting to the facsimile machine of the other party via the public line 113. Furthermore, for sending / receiving E-mail and browsing Web pages, after connecting to the ISP via the public line 113, T
CP / IP (Transmission Control Protocol / Interne
t Protocol) Transmission and reception of communication data according to the protocol.

FIG. 2 is a diagram conceptually showing the structure of the voice dictionary stored in the voice dictionary memory 104.

As shown in the figure, the voice dictionary memory 104 classifies and stores standard voice patterns into a plurality of voice dictionaries according to their types. In the figure, the notation between "" indicates a standard voice pattern, and the notation between "" indicates a character or an operation command corresponding to each standard voice.

The dictionary A indicated by 201 is a kana standard voice pattern, and stores standard voice patterns of Japanese syllabary such as "A", "I", "U". This dictionary A is mainly used for voice recognition when voice inputting the connection destination abbreviation.

The dictionary B indicated by 202 is a standard voice pattern for numbers, and stores standard voice data of numbers such as "1", "2", "3". This dictionary B is mainly used for voice recognition when the telephone number of the connection destination is input by voice.

The dictionary C indicated by 203 is a standard voice pattern for the alphabet, and the alphabet "a",
Standard voice patterns such as “b” and “c” and symbols “@”, “/”, “:” are stored. This dictionary C is mainly used for voice recognition when voice inputting an Internet address.

A dictionary D indicated by 204 is a standard voice pattern for operation commands, and includes device control commands such as "start", "stop", "register", "telephone", and "FA".
A standard voice pattern of a command for selecting a device function such as "X" or "E-mail" is stored.

As described above, in this embodiment, the dictionaries A, B,
Each C is used as a connection destination input dictionary for inputting information for identifying a connection destination, and the dictionary D is used as an operation command dictionary. In the communication device of the present embodiment, the voice dictionary is classified into four types as described above, but different classification or more detailed classification may be performed depending on the purpose and use of voice recognition.

The configuration of the communication device according to this embodiment has been described above.

Next, the operation of the communication apparatus of the present invention will be described with reference to the flow charts of FIGS.

FIG. 3 is a flowchart showing the operation of the communication device 100 when the telephone number is input by the voice of the user. Programs corresponding to these flowcharts are stored in a ROM (not shown) included in the control unit 114 and executed by the control unit 114.

In the initial state of the communication device 100, its operation mode is set to the operation standby mode for waiting for the input of an operation command. In the operation standby mode, the standard voice pattern 204 for operation command is selected by the dictionary selection unit 103 (step S301), and if any operation is performed by the user's voice, the process proceeds to the next step (step S302). That is, the communication device 100
The operation standby mode is shifted to the input mode in which the information specifying the connection destination can be input. The input modes in this embodiment include an address book registration mode, an address book search mode, an e-mail & Internet mode, and a TEL & FAX mode, as shown below.

Here, for example, when the user inputs the voice "Address butterfly" + "Touroku" and the voice is recognized by the voice recognition unit 102 as the operation command "Address book" + "Registration", communication is performed. The device 100 shifts to the flow of FIG. 5 as the address book registration mode under the control of the control unit 114 (step S303).

Also, "address butterfly" + "kensaku"
6 is input and recognized by the voice recognition unit 102 as an operation command of “address book” + “search”, the communication device 100 shifts to the flow of FIG. 6 as the address book search mode under the control of the control unit 114. (Step S304).

Further, when the voice of “mail” or “Internet” is input and recognized by the voice recognition unit 102 as an operation command of “Email” or “Internet”, the communication device 100 is controlled by the control unit 114. The flow shifts to the flow of FIG. 4 as the e-mail & Internet mode (step S305).

If the operation command recognized in step S302 is "telephone" or "FAX",
It is considered that the user is making a call or a facsimile transmission (step S305), and the process proceeds to a voice telephone number input process.

In order to input the telephone number by voice, the control unit 114 selects the standard voice pattern for operation command 204 and the standard voice pattern for numbers 202 as the voice dictionary in the dictionary selection unit 103 (step S306). It is displayed on the display unit 106 that the number can be recognized by voice (step S307). Next, the telephone number input by the user's voice is recognized.

The voice recognition unit 102 recognizes the input voice character by character (step S308), and displays the recognized number on the display unit 106 (step S309). This process is repeated until the voice "start" indicating the start of calling is input (steps S308 to S31).
0).

For example, the telephone number is 03-1234.
When the user voice-inputs the number 5678, the user first inputs the voice "zero". The voice recognition unit 102 recognizes this voice as a number “0” and displays <0> on the display unit 106. Then "Sun"
Is input, the same voice recognition processing is performed, and <03> is displayed on the display unit 103. This process is repeated until <0312345678> is displayed on the display unit 103 (steps S308 to S310).

In step S310, the voice "start" is input from the user, and when the voice recognition unit 102 recognizes this voice as the operation command "start", the telephone call process is subsequently performed. Control unit 114
Sends out the telephone number recognized in steps S308 to S310 to the public line 113 via the communication control unit 112 (step S311), and connects to the destination telephone or facsimile machine.

Next, it is determined whether the voice input by the user in step S302 is a command indicating a "telephone" operation or a command indicating a "facsimile" operation (step S312). here,
If the voice input by the user is a command indicating a “telephone” operation, the call unit 111 is set to the public line 113.
To make a call (step S313). On the other hand, when the voice input by the user is a command indicating a “facsimile” operation, the image data read by the reading unit 107 is converted into G3 facsimile data, and the facsimile of the other party is transmitted via the public line 113. It is transmitted to the device (step S314).

After the call or the facsimile transmission is completed, the communication device 100 returns to the initial state (operation standby mode),
Standard voice pattern 2 for operation command in dictionary selection unit 103
04 is selected (step S301), and the next operation by the user is awaited.

FIG. 4 is a flow chart showing an operation when inputting an e-mail address or an internet address by the voice of the user. As mentioned earlier, Figure 3
If it is determined to be the e-mail & Internet mode in step S305 of step S305, the e-mail transmission or W
Since it is assumed that the eb page is browsed, the process shifts to a voice address input process.

First, the control unit 114 uses the standard voice pattern 202 for numbers, the standard voice pattern 203 for alphabets, and the standard voice pattern 20 for operation commands in the dictionary selection unit 103 to input an address by voice.
4 is selected as a voice dictionary (step S401), and it is displayed on the display unit 106 that voice recognition of numbers and alphabets is possible (step S402).

Next, the address input by the voice of the user is recognized. The input voice is converted into the voice recognition unit 102.
In step S403, the characters are recognized one by one, and the recognized characters are displayed on the display unit 106 (step S40).
4). This process is repeated until a voice "START" indicating a start operation command is input (step S
403 to step S405).

For example, when the e-mail address abc@123.co.jp is input as the address, the user first inputs the voice "A". The voice recognition unit 102 recognizes this voice as a number “a” and displays <a> on the display unit 106. Next, when the voice "BEE" is input, the same voice recognition processing is performed, and <ab> is displayed on the display unit 103. This process is performed by the display unit 10
This is repeated until <abc@123.co.jp> is displayed in step 3 (step S403 to step S405).

The user inputs the voice "start", and this voice is recognized by the voice recognition unit 102 in step S405.
When it is recognized that the operation command is “start”, the ISP telephone number registered in advance is sent to the public line 113 via the communication control unit 112 to establish connection with the ISP (step S406). .

Next, whether to send and receive e-mail,
Whether to browse the page is determined in step S302 described above.
Judgment based on the operation command in step S40 (step S40)
7). If the operation command indicating "Internet" is recognized in step S302, step S
The Internet address input in steps 403 to S405 is sent to the ISP (step S408), and the information of the corresponding web page is downloaded via the ISP (step S409). The downloaded information is temporarily stored in the data memory 109, and the contents are displayed on the display unit 1.
The web page is displayed at 06 or printed at the recording unit 108 (step S410) to browse the web page.

On the other hand, if the operation command indicating "Email" is recognized in step S302, step S302
The email address input in steps 403 to S405 is inserted into the destination field of the email created in advance (step S411), and the SMTP (Simple Mail Transfer Proto) is entered.
col) protocol, the e-mail information is transmitted to the ISP (step S412).

The communication device 100 sends e-mail or W
After the browsing of the eb page is completed, the process returns to the initial state (operation standby mode), the standard voice pattern 204 for operation command is selected by the dictionary selection unit 103 (step S301), and the next operation by the user is awaited.

It should be noted that the voice recognition function can also be used in creating the body of the email to be sent. In this case, the corresponding voice dictionary may be switched each time according to the type of character to be input, and the document may be created by voice input.

FIG. 5 is a flow chart showing the operation when the address book is registered by the voice of the user. As described above, when the address book registration mode is determined in step S303 of FIG. 3, it is considered that the address book registration is performed, and the process proceeds to the process of inputting the connection destination abbreviation and the connection destination address by the voice of the user.

First, the dictionary selecting section 103 selects the kana standard voice pattern 201 and the operation command standard voice pattern 204 as a voice dictionary (step S50).
1) The display unit 106 displays that kana voice recognition is possible (step S502).

Next, the voice recognition unit 102 recognizes the voice of the connection destination abbreviation inputted by the user character by character (step S503) and displays the recognized character on the display unit 106 (step S504). This process is repeated until a voice "SET" indicating the operation command for determination is input (steps S503 to S505).

For example, when the abbreviation "Yamada" is input as the abbreviation, the user first inputs the voice "YA". The voice is recognized by the voice recognition unit 102 as “ya”.
And displays <or> on the display unit 106. next,
When the voice "MA" is input, the same voice recognition process is performed, and "Yama" is displayed on the display unit 103. This process is repeated until <Yamada> is displayed on the display unit 103 (steps S503 to S505).

In step S505, when the user inputs the voice "set" and the voice recognition unit 102 recognizes that it is the "determine" command, the dictionary selection unit 103 then inputs the voice of the connection destination address. Selects the standard voice pattern for numbers 202, the standard voice pattern for alphabets 203, and the standard voice pattern for operation commands 204 as a voice dictionary (step S50).
6) The display unit 106 displays that voice recognition of numbers and alphabets is possible (step S507).

Next, the voice recognition unit 102 recognizes the voice of the connection destination address input by the user character by character (step S508) and displays the recognized character on the display unit 106 (step 509). This process is repeated until a voice "SET" indicating a determination operation command is input (steps S508 to S510). The input destination address may be a telephone number, an email address, or an internet address.

In step S510, the user inputs a voice "set", and when the voice recognition unit 102 recognizes this voice as a "decide" command, the connection input in the processes of steps S503 to S505. The connection destination address input in the processing of steps S508 to S510 is added to the abbreviation and registered in the address book 110.

After the address book registration is completed, the communication device 100 returns to the initial state, and the dictionary selection unit 103 selects the standard voice pattern 204 for operation command (step S30).
1) Wait for the next operation by the user.

In the address book registration of FIG. 5, the abbreviations are registered using hiragana characters, but a kanji conversion ROM is installed in the communication device 100, and hiragana characters are converted to kanji characters such as [Yamada] → [Yamada]. Then, it may be registered in the address book. In this case, a standard voice pattern “Henkan” indicating the operation command of “Kanji conversion” may be prepared as the standard voice pattern for operation command, and the Kanji conversion may be performed by the voice of the user.

FIG. 6 is a flow chart showing the operation when the address book is searched by the voice of the user. As described above, when the address book search mode is determined in step S304 of FIG. 3, it is considered that the address book search is performed, and the process proceeds to the process of inputting the connection destination abbreviation by the voice of the user.

First, the dictionary selection unit 103 selects the kana standard voice pattern 201 and the operation command standard voice pattern 204 as a voice dictionary (step S60).
1) The display unit 106 displays that voice recognition of a kana is possible (step S602).

Next, the voice recognition unit 102 recognizes the voice of the connection destination abbreviation input by the user character by character (step S603), and displays the recognized characters on the display unit 106 (step S604). This process is repeated until the voice "SET" indicating the operation command for determination is input (steps S603 to S605).

In step S605, the user inputs the voice "set", and when this voice is recognized by the voice recognition unit 102 as the "decide" command, the address book search is performed subsequently. The input connection abbreviation and the connection destination abbreviation registered in the address book 110 are searched for (step S606). If they match, the abbreviation and address are displayed on the display unit 106 (step S607). .

Next, when the user inputs the voice "start" and this voice is recognized by the voice recognition unit 102 as the "start" command (step S6).
08), the processing corresponding to the connection destination address retrieved from the address book 110 is executed. If the retrieved address is a telephone number, the number is sent to the public line 113 to make a call or send a facsimile. If the retrieved address is an e-mail address or an internet address, connection is made with the ISP to send e-mail or browse a web page (step S6).
09).

After the address book search is completed, the communication device 100 returns to the initial state, and the dictionary selection unit 103 selects the standard voice pattern 204 for operation command (step S30).
1) Wait for the next operation by the user.

As described above, according to the communication device of this embodiment, the voice dictionary used for recognition processing is limited depending on the operation mode of the communication device. Therefore, the standard voice pattern to be subjected to the pattern matching is smaller than that in the conventional case, so that the load of the voice recognition processing can be reduced and the processing speed can be further increased.

Further, since a plurality of voice dictionaries in which standard voice patterns are classified according to the type of voice are prepared and the optimum voice dictionary is selected according to the operation mode of the communication device, an incorrect standard voice pattern is selected in voice recognition. Less likely to improve recognition rate (eg,
When the communication device is in the telephone number input mode, the standard voice pattern for numbers is selected as the voice dictionary and the other standard voice patterns are not used, so that characters other than numbers are not output as the recognition result).

Further, since the operation mode of the communication device is displayed on the display section of the device, the user can easily know the type of voice to be input.

Further, since the address book can be registered / searched by the voice of the user, even in a communication device in which the keys of the operation section are limited as compared with the keyboard of a personal computer, it is easy to input the e-mail address or the internet address. Therefore, it is possible to greatly improve the operability of the communication device.

(Embodiment 2) In this embodiment, a communication device capable of discriminating which function is desired to be used only by inputting the other party's abbreviation is described. It should be noted that in the present embodiment, the browsing of the Web follows the HTTP protocol, and thus the URL is used as the Internet address.

FIG. 7 is a block diagram of the communication device in this embodiment.

In the figure, 701 is a CPU as a control unit for controlling the operation of the entire apparatus and data processing, and 702 is a CP.
It is a ROM for storing various programs executed by the U 701 and fixed data. Also, the ROM 702
A phoneme model used in the speech recognition processing is also stored in (details will be described later). A RAM 703 is used as a work memory for the CPU 701 and a buffer memory for voice, image data, and the like.

Reference numeral 704 denotes a contact type color image sensor CIS (Conta) for reading image data line by line.
ct Image Sensor) has a built-in red (R), green (G), and blue (B) LED array, and when a read timing signal is input, a video signal is output in synchronization with the timing signal. .

Reference numeral 705 outputs a read timing signal to the CIS 104, A / D-converts the video signal output by the CIS 104, and then performs shading correction and luminance /
This is a read image processing unit that performs gamma conversion of density to convert it into digital multi-value / binary data, and performs image enlargement / reduction and resolution conversion. The image data converted into digital data is stored in the RAM 703.

Reference numeral 706 carries out necessary image processing such as binarization processing for converting multi-valued image data into a binary image and image enlargement / reduction, and the resolution of the image data is changed to the printer 70.
Resolution conversion to convert to a resolution of 360 dpi of 7
A recording image processing unit that transfers print data to a color printer. Reference numeral 707 denotes a known color printer of, for example, an ink discharge method as a printer unit of this apparatus,
While moving the cartridge in which the ink tank and the ink discharge unit are integrated in the main scanning direction of the recording paper, 360 d
Binary data is recorded on a recording paper with a recording resolution of pi.

Reference numeral 708 designates a FAX modem for modulating / demodulating data for performing facsimile communication in the well-known G3 mode and the like, and a well-known DATA modem for utilizing the Internet service, and 709 denotes an analog telephone line 710.
And known NCU for interfacing with this device
Is. Further, a handset 716 for a call is connected to the NCU 709, and when a call is made, the off-hook of the handset is detected, and the telephone line 710 and the handset 7 are detected.
16 are connected.

Reference numeral 711 is a dial operation or various operation instructions,
Mode setting, data registration / search / selection,
Consists of a key input unit consisting of an Internet button for using Internet services and a mail button for using E-mail, and a liquid crystal display unit and LEDs for displaying the status of the device, various data and search results for voice recognition, which will be described later. This is the operation panel.

Reference numeral 712 is a speaker for outputting various alarm sounds, ringing sounds and voice signals on the line, and 713 is a microphone for inputting voice. It is desirable to use a directional microphone for this microphone 713 in order to reduce the mixing of noise components as much as possible.

Reference numeral 714 is a non-volatile memory in which data is not lost even when the main power is cut off, and stores registration data (various soft switches, ID information such as user's telephone number and abbreviated name) necessary for the operation of this apparatus. ing. In addition, as will be described later, destination information registered by the user (specifically, a telephone number, a facsimile number, a mail address or a UR).
L) is stored in association with the other party's abbreviation.

Reference numeral 715 denotes a sensor section composed of a recording paper width sensor, a recording paper presence sensor, a document width sensor and a document presence sensor, and detects the condition of the document and the recording paper under the control of the CPU 701.

Reference numeral 717 is a voice recognition processing unit, the detailed structure of which will be described later.

Each functional block is connected via an address / data bus 720.

FIG. 8 is a diagram showing a block configuration and a signal flow relating to the voice recognition processing in this embodiment.
It should be noted that the signal flow is for the purpose of explanation, and the actual signal may be connected via the address / data bus 720 even if it is directly connected on the flow.

As shown in the figure, the voice recognition processing section 717 includes a voice analysis section 83, a phoneme recognition section 84, and a word determination section 85.
It is composed of.

The voice analysis unit 83 divides the voice signal input from the microphone 713 into predetermined minute frames and performs predetermined filtering and signal calculation processing for the purpose of removing noise components. In addition, in order to speed up the signal processing, DSP
Is used. The voice data divided into minute frames by the voice analysis unit 83 is input to the phoneme recognition unit 84.

In the phoneme recognition unit 84, a phoneme model 702a prepared beforehand (stored in the ROM 702).
Based on, the frequency spectrum of each minute frame is compared and collated including the time change, and the phoneme is specified.

The word determining unit 85 compares and collates the specified phoneme model with a previously registered word model, thereby calculating the matching frequency between the two and specifying the word, and outputting the result. This output result is collated with each partner abbreviation in the partner abbreviation table 81 stored in the non-volatile memory 714, and the partner abbreviation with the highest degree of similarity is specified. Then, according to the operating state of the device, the destination information corresponding to the specified abbreviation of the destination is selected from the destination information table 82 stored in association with the abbreviation of the destination. The destination information table 82 is, as shown in the figure, a telephone number table, a FAX number table, a mail address table, a U number table.
Contains the RL table.

Next, the operation of the communication apparatus according to this embodiment will be described in detail with reference to the flowchart of FIG.

First, in the standby state, the state of the handset 716 is determined (step S1), and if off-hook is detected, the process proceeds to step S5, and it is determined whether or not there is a voice input from the microphone 713. If there is a voice input, the voice recognition processing unit 717 identifies the phoneme in step S6, and the nonvolatile memory 714 is identified from the recognition result.
Specify the abbreviation of the other party registered in. When the destination abbreviation is specified, in step S7, the nonvolatile memory 714
The telephone number table registered in is searched for the telephone number associated with the destination abbreviation, the result is displayed on the display unit of the operation panel 711 as the destination abbreviation / telephone number, and the speaker 712 displays the destination. Announce the abbreviation to the operator by voice.

Then, in step S8, the searched telephone number is automatically dialed, and in step S9, the telephone is returned to the standby state after the call. If there is no voice input of the other party's abbreviation in step S5, the operator dials the call manually in step S8 to make a call.

Next, when it is determined in step S1 that the hook is on-hook, the process proceeds to step S2, and it is determined whether or not there is a document. When the sensor section 715 detects the presence of a document, it is determined in step S10 whether or not there is a voice input from the microphone 713. If there is voice input, step S1
1, the speech recognition processing unit 717 identifies phonemes,
From the recognition result, the other party's abbreviation registered in the non-volatile memory 714 is specified.

When the other party's abbreviation is specified, step S1
FA registered in the non-volatile memory 714 in No. 2
Referring to the X number table, the FAX number associated with the destination abbreviation is searched, and the result is displayed as the destination abbreviation / FAX number on the display unit of the operation panel 711, and the speaker 7 is displayed.
From 12 the voice of the other party is announced to the operator.

Then, in step S13, the retrieved F
Automatically dials the AX number, and in step S14 FA
As soon as X transmission is completed, it returns to the standby state again.

If there is no voice input of the other party's abbreviation in step S10, the process proceeds to step S24,
Whether or not the copy key is pressed is determined. If the copy key is pressed, the copy operation is performed in step S25 and the operation returns to the standby state. If the copy key is not pressed, the process proceeds to step S13, and the fax transmission is performed after the dial call is manually made by the operator.

If it is determined in step S1 that the document is on-hook and that there is no document in step S2, the flow advances to step S3 to determine whether or not the mail key for creating an email has been pressed. If the mail key is pressed, the process proceeds to step S15, and the microphone 71
The presence or absence of voice input from 3 is determined.

If there is voice input, step S16.
Then, the speech recognition processing unit 717 identifies the phoneme, and identifies the partner abbreviation registered in the non-volatile memory 714 from the recognition result. When the destination abbreviation is specified, in step S17, the email address table registered in the non-volatile memory 714 is referred to, the email address associated with the destination abbreviation is searched, and the result is displayed on the operation panel 711. The other party abbreviation / email address is displayed on the display unit, and the other party abbreviation is announced to the operator by voice from the speaker 712.

After that, in step S18, the address is automatically entered in the destination field (To :), and in step S1
After the mail text is created in 9, the mail is sent in step S20, and the standby state is restored again.

When the voice of the recipient's abbreviation is not input in step S15, the mail address is manually input by the operator through the ten-key pad, and the mail text is created, and then the mail is transmitted. It will be.

Next, it is determined in step S1 that the document is on-hook, and in step S2 that there is no document,
If it is determined in step S3 that the mail key has not been pressed, the process proceeds to step S4, and it is determined whether or not there is voice input from the microphone 713. If there is a voice input, the process proceeds to step S21, the phoneme is identified by the voice recognition processing unit 717, and the destination abbreviation registered in the non-volatile memory 714 is identified from the recognition result.

When the destination abbreviation is specified, the URL table registered in the non-volatile memory 714 is referred to in S22, the URL associated with the destination abbreviation is searched, and the result is displayed on the operation panel 711. The other party's abbreviation / URL is displayed on the display unit, and the other party's abbreviation is announced by voice from the speaker 712. After that, the process proceeds to step S23, and the searched U is searched by using the Internet service.
The RL is automatically accessed, and the home page information designated by the URL is displayed on the display unit of the operation panel 711.

According to the communication device of the present embodiment described above, in the image communication device having the voice recognition means, voice recognition is performed on the input voice, and the result is used as the abbreviations of the destinations stored in the table. When collating, by limiting the reference range of the table according to the operation state of the device, it is possible to efficiently search even if a large number of destination information is registered.

Further, according to the present embodiment, in a multi-function (including a call function, a facsimile communication function, an e-mail transmission / reception function) type communication device, the operating state of the communication device is determined in advance to determine which one. Since the function to be used is determined, it becomes possible for the user to omit the operation of specifying which function to use.

(Third Embodiment) In the above-described first embodiment, in the address book search mode, the kana standard voice pattern 201 and the operation command standard voice pattern 204 are selected as the voice dictionary, and the voice input connection is made. Although the voice recognition of the destination (destination) abbreviation was performed, a standard voice pattern for inputting the destination abbreviation is prepared in advance,
This may be used to perform voice recognition of the other party's abbreviation. The details will be described below.

Since the configuration of the communication device in this embodiment is the same as that in FIG. 1, its description is omitted.

In the present embodiment, the voice pattern of each destination abbreviation registered in advance in the address book 110 in the voice dictionary memory 104 shown in FIG. 2 (hereinafter referred to as a word model).
Is further stored in the dictionary. The word model indicating the word indicating the other party abbreviation may be generated from the abbreviation when the abbreviation and the address are registered in the address book 110 in S511 of the address book registration mode in FIG. In addition, in S503 to S504, the method of displaying characters from the voice input one by one is adopted.
The input voice may be directly registered in the address book 110 as a word model indicating the abbreviation of the other party. In this way, the voice dictionary of the word model representing the abbreviation of the address book 110 is generated in advance.

The outline of the operation of the communication device in this embodiment is the same as the flow of FIG. 3 shown in the first embodiment. However, when the address book search mode is selected in step S304 of FIG. 3, the flow shifts to the flow shown in FIG. 10 instead of shifting to the flow shown in FIG. The flow of FIG. 10 will be described below.

When the address book search mode is selected in step 304 of FIG. 3, first, the dictionary selection section 103 selects the word model of the other party's abbreviation and the standard voice pattern 204 for operation command as a voice dictionary (step S100).
1) The display unit 106 displays that the voice recognition of the other party's abbreviation is possible (step S1002).

Here, all the other party's abbreviations in the address book 110 are subject to voice recognition.

Next, it is judged whether or not the user inputs a voice corresponding to the other party's abbreviation (for example, "Yamada" voice) (step S1003), and if there is a voice input, it is subsequently input as an address book search. The generated voice is compared with the dictionary of the registered destination abbreviation word model (step S1004). It is determined whether or not the input voice matches the word model of the connection destination abbreviation registered in the address book 110 (step S1005),
If they match, the abbreviation and address are displayed on the display unit 106 (step S1006). As a result, the voice dictionary to be recognized is appropriately limited, so that a quick and accurate address book search can be performed.

Next, when the user inputs the voice "start" (or presses the start key or the like on the operation unit 105), the voice is recognized by the voice recognition unit 10.
In step 2, the command is recognized as a "start" command (step S1007). When the start command is recognized, the processing corresponding to the connection destination address retrieved from the address book 100 is executed. That is, when the retrieved address is a telephone number, that number is sent to the public line 113 to make a call or a facsimile transmission. If the retrieved address is an e-mail address or an internet address, connection with the ISP is made, and e-mail is transmitted or a web page is browsed (step S10).
08). On the other hand, if it is determined in step 1005 that the input voice does not match the word model of the connection destination abbreviation registered in the address book 110, a display indicating that the voice recognition has failed is displayed (step S100
9).

On the other hand, before voice input in step S1003, it is determined whether or not the Internet key on the operation unit 105 is pressed (step S1010). If it is determined that the Internet key is pressed, the word model of the operation command, E-mail address and We in address book
b Only the word model of the other party's abbreviation in which the internet address of the browser is registered is selected as the voice dictionary (step S1011). This further limits the voice dictionary, allowing for quick and accurate address book searches.

Here, it is determined whether or not the Internet key is pressed, but step S305 in FIG.
As described above, the voice "Internet" may be input as the operation command.

If the "mail key" is pressed as shown in step S3 of FIG. 9 of the second embodiment, E
Only the word model of the other party abbreviation for which the -mail address is registered may be selected as the voice dictionary.

When the voice dictionary is selected in step S1011, the process of step S1003 is performed.

If it is determined in step S1010 that the Internet key has not been pressed, then the call unit 111
It is determined whether the telephone is off-hook (step S1012). If it is judged that the phone is off-hook, it is judged that it will be used for telephone or FAX transmission, and only the word model of the operation command and the word model of the abbreviation of the recipient whose telephone number for the telephone or FAX is registered in the address book are used. , As a voice dictionary (step S101)
3). This further limits the voice dictionary, allowing for quick and accurate address book searches. After the voice dictionary is selected, the process of step S1003 is performed.

In step S1012, if the telephone of the call section 111 is on-hook, the process proceeds to step S1014. In step S1014, a sensor (not shown) (may be the sensor unit 715 of the second embodiment) determines whether or not there is a document in the reading unit 107. If there is a document, the word model of the operation command and the fax phone are used. Only the word model of the other party's abbreviation for which the number is registered is selected as the voice dictionary (step S1015). When the voice dictionary is selected, the process of step S1003 is performed.

Here, when there is a manuscript, only the word model of the other party's abbreviation in which the FAX telephone number is registered is selected as the voice dictionary, but the present invention is not limited to this, and the read image is not used. When there is a function of transmitting as an attached file of E-mail, the word model of the other party abbreviation in which the E-mail address is registered may be added to the voice dictionary. When the function of sending the read image to the Internet address of the Web browser is provided, the word model of the other party's abbreviation may be added to the voice dictionary.

Also with the communication device of the present embodiment described above, the voice dictionary used for the recognition processing is limited depending on the operation mode of the communication device, and it is possible to achieve the same effect as in the above-described first embodiment.

[0132]

Other Embodiments Although the embodiments of the present invention have been described in detail above, the present invention can be applied to a system including a plurality of devices (eg, host computer, interface device, reader, printer, etc.). It may be applied to an apparatus (for example, a copying machine, a facsimile apparatus, etc.) which is composed of one device.

The present invention supplies a software program for realizing the functions of the above-described embodiments to a system or apparatus directly or remotely, and a computer of the system or apparatus reads and executes the supplied program. Including the case where it is achieved by

Therefore, the program code itself installed in a computer to implement the functional processing of the present invention by the computer also implements the present invention. That is, the claims of the present invention include the computer program itself for realizing the functional processing of the present invention.

In this case, the program may take any form such as an object code, a program executed by an interpreter, or script data supplied to an OS as long as it has the function of the program.

Storage media for supplying the program include, for example, floppy (registered trademark) disks, optical disks (CD-ROM, CD-R, CD-RW, DVD, etc.), magneto-optical disks, magnetic tapes, memory cards. Etc.

As another method of supplying the program,
A mode in which the program of the present invention is acquired by file transfer via the Internet is also included.

[0138] The program of the present invention is encrypted to be a CD.
-Stored in a storage medium such as a ROM and distributed to users, users who have met prescribed conditions are allowed to acquire key information for decryption via the Internet, and the key information is used for encryption. It is also possible to execute the program and install the program on a computer to realize it.

The computer executes the read program to realize the functions of the above-described embodiments, and the OS or the like running on the computer executes the actual processing based on the instructions of the program. The functions of the above-described embodiments may be realized by performing a part or all of the processing.

Furthermore, after the program read from the storage medium is written in the memory provided in the function expansion board inserted into the computer or the function expansion unit connected to the computer, based on the instructions of the program,
CPU provided on the function expansion board or function expansion unit
Etc. perform a part or all of the actual processing, and the functions of the above-described embodiments are also realized by the processing.

[0141]

According to the present invention, in a communication device equipped with a voice recognition means, it is possible to reduce the load of voice recognition processing, speed up the processing, and improve the voice recognition accuracy. Further, according to the present invention, in a multi-function communication device (including, for example, a call function, a facsimile communication function, an e-mail transmission / reception function, etc.), it is possible to determine which function to use only by voice inputting the abbreviation of the other party. Therefore, it is possible to simplify the operation procedure.

[Brief description of drawings]

FIG. 1 is a block configuration diagram of a communication device according to a first embodiment.

FIG. 2 is a diagram conceptually showing the structure of a voice dictionary stored in a voice dictionary memory according to the first embodiment.

FIG. 3 is a flowchart showing an operation of the communication device according to the first embodiment when a telephone number is input by a voice of a user.

FIG. 4 is a flowchart showing an operation of the communication device according to the first embodiment when inputting an e-mail address or an internet address by voice of a user.

FIG. 5 is a flowchart showing an operation of the communication device according to the first embodiment when registering an address book by a voice of a user.

FIG. 6 is a flowchart illustrating an operation of the communication device according to the first exemplary embodiment when an address book search is performed by a user's voice.

FIG. 7 is a block configuration diagram of a communication device according to a second embodiment.

FIG. 8 is a diagram showing a block configuration and a signal flow relating to a voice recognition process in a second embodiment.

FIG. 9 is a flowchart illustrating an operation of the communication device according to the second exemplary embodiment.

FIG. 10 is a flowchart showing an operation of the communication device according to the third exemplary embodiment when an address book search is performed by a user's voice.

─────────────────────────────────────────────────── ─── Continuation of front page (51) Int.Cl. ⁷ Identification code FI theme code (reference) H04M 1/274 H04M 11/00 302 1/56 G10L 3/00 551A 11/00 302 521V F term (reference) 5D015 KK01 5K027 BB01 EE04 FF22 HH20 5K036 BB01 DD17 FF06 JJ02 JJ13 5K101 NN07 NN18 NN21

Claims (23)

[Claims] 1. A communication device comprising voice recognition means for recognizing an input voice, which can be controlled according to the output of the voice recognition means, and storage means for storing a plurality of types of voice recognition dictionaries. A communication device comprising: a selection unit that selects a voice recognition dictionary according to an operation mode of the communication device from the storage unit and supplies the dictionary to the voice recognition unit. 2. The plurality of voice recognition dictionaries include an operation command dictionary in which standard voice patterns for operation commands for the communication device are stored, and the selection unit has the operation mode input of the operation command. The communication device according to claim 1, wherein the operation command dictionary is selected in a standby operation standby mode. 3. The plurality of voice recognition dictionaries further includes a connection destination input dictionary that stores a standard voice pattern for inputting information for identifying a connection destination, and the selection unit further includes the operation mode. 3. The communication device according to claim 2, wherein the connection destination input dictionary is additionally selected when is in an input mode of waiting for input of information specifying a connection destination. 4. The plurality of voice recognition dictionaries further includes a connection destination abbreviation input dictionary that stores a standard voice pattern for inputting information for identifying a connection destination abbreviation, and the selection unit further includes: The communication device according to claim 2 or 3, wherein the connection destination abbreviation input dictionary is additionally selected when the operation mode is an input mode for waiting for input of information specifying the connection destination abbreviation. . 5. The display device further comprises display means for displaying the types of voices that can be recognized by voice, in accordance with the voice recognition dictionary supplied to the voice recognition means by the selecting means. 4. The communication device according to any one of 4 to 4. 6. A method of controlling a communication device, comprising: a storage unit for storing a plurality of types of voice recognition dictionaries, the determination step comprising: determining an operation mode of the communication device; A selection step of selecting a voice recognition dictionary corresponding to the determined operation mode and supplying the voice recognition means, a voice input step of inputting voice, and a voice input step of inputting the voice in the selection step. A method of controlling a communication device, comprising: a voice recognition step of performing a voice recognition process using the voice recognition dictionary; and a step of executing a process in the operation mode based on an output of the voice recognition process. . 7. The plurality of voice recognition dictionaries include a manipulation command dictionary in which standard voice patterns for manipulation commands for the communication device are stored, and in the determining step, the operation mode indicates that the manipulation command is input. 7. The control method of the communication device according to claim 6, wherein when it is determined that the operation standby mode for waiting is selected, the operation command dictionary is selected in the selection step. 8. The plurality of voice recognition dictionaries further includes a connection destination input dictionary that stores a standard voice pattern for inputting information for identifying a connection destination, and in the determining step, the operation mode is connection. 8. The connection destination input dictionary is additionally selected in the selection step when it is determined that the input mode is to wait for the input of information specifying the destination, and the connection destination input dictionary is additionally selected. Method for controlling communication device of the present invention. 9. The plurality of voice recognition dictionaries further includes a connection destination abbreviation input dictionary storing a standard voice pattern for inputting information for identifying a connection destination abbreviation, and in the determining step, the operation mode is set. Is determined to be an input mode for waiting for the input of information specifying a connection destination abbreviation, the connection destination abbreviation input dictionary is additionally selected in the selecting step. Item 9. A method for controlling a communication device according to item 7 or 8. 10. A program for controlling a communication device comprising a storage means for storing a plurality of types of speech recognition dictionaries, the program code of a determination step for determining an operation mode of the communication device, and the storage means. From the selection, the voice recognition dictionary corresponding to the operation mode determined in the determination step is selected, and the program code of the selection step to be supplied to the voice recognition means, and the program code of the voice input step of inputting voice, The program in the voice recognition step that performs voice recognition processing using the voice recognition dictionary supplied in the selection step on the generated voice, and the processing in the operation mode is executed based on the output of the voice recognition processing. And a program code of a step, and a program comprising: 11. A storage medium storing a program for controlling a communication device, comprising a storage means for storing a plurality of types of speech recognition dictionaries, the program code of a determination step of determining an operation mode of the communication device. A program code of a selection step for selecting a voice recognition dictionary corresponding to the operation mode determined in the determination step from the storage means, and supplying the voice recognition means to the voice recognition means; and a voice input step of inputting voice. A program code, a program code of a voice recognition step of performing voice recognition processing on the inputted voice using the voice recognition dictionary supplied in the selection step, and the operation mode based on an output of the voice recognition processing. And a program code of steps for executing the process in Medium. 12. A communication device comprising: a call means for making a call via a telephone line, a facsimile communication means for making a facsimile communication via a telephone line, and a data communication means for making a data communication via the Internet. Storage means describing the correspondence between the connection destination abbreviation and the information specifying the connection destination, a voice input means for inputting a voice, a voice recognition means for recognizing the input voice, and each of the communication means. A determining means for determining which one of the means to be used, a collating means for collating the output of the voice recognizing means with each connection destination abbreviation in the storage means, and selecting the connection destination abbreviation according to the comparison result. And, the collation unit includes a control unit that limits the collation candidates of each connection destination abbreviation in the storage unit according to the determination result of the determination unit. Characterized communication device. 13. The storage means stores a plurality of tables in which the correspondences are classified and described for each of the communication means, and the control means responds to the determination result among the plurality of tables. The communication device according to claim 12, wherein the matching candidates are limited by selecting a table. 14. The deciding means decides to use the call means when the communication device is off-hook and a voice is inputted by the voice input means. The communication device described. 15. The determination means determines to use the facsimile communication means when a voice is input by the voice input means in a state where a facsimile transmission original is set in the communication means. The communication device according to any one of claims 12 to 14. 16. The determination unit determines that the data communication unit is to be used when the communication unit is on-hook and a voice is input by the voice input unit. The communication device according to any one of 1 to. 17. The information for identifying the connection destination includes a telephone number and a facsimile number, and further includes an electronic mail address and / or an internet address, according to any one of claims 12 to 16. Communication device. 18. A telephone calling means for calling a telephone number corresponding to the connection destination abbreviation selected by the collating means when the judging means determines that the calling means is to be used. 18. The communication device according to claim 17, wherein the communication device is a communication device. 19. A facsimile calling means for calling a facsimile number corresponding to the connection destination abbreviation selected by the collating means when the judging means determines to use the facsimile communication means. The communication device according to claim 17 or 18, characterized in that. 20. Address input means for inputting an electronic mail address and / or an internet address corresponding to the connection destination abbreviation selected by the collating means when the judging means determines to use the data communication means. 18. The method according to claim 17, further comprising:
9. The communication device according to any one of 9 to 9. 21. A storage unit describing a correspondence relationship between a connection destination abbreviation and information specifying the connection destination, a call unit for making a call via a telephone line, and a facsimile communication for making a facsimile communication via a telephone line. A voice input step of inputting a voice, a voice recognition step of recognizing the input voice, comprising: A determination step of determining which one of the communication means is used, the recognition result obtained in the voice recognition step and each connection destination abbreviation in the storage means are collated, and connection is made according to the collation result. A collating step of selecting an abbreviation, and the collating step according to the determination result obtained in the determining step; The communication control method characterized by comprising the control step of limiting the matching candidates for the destination abbreviation. 22. Storage means describing a correspondence relationship between a connection destination abbreviation and information for identifying the connection destination, a communication means for making a call via a telephone line, and a facsimile communication for making a facsimile communication via a telephone line. And a data communication means for performing data communication via the Internet, the program being for controlling a communication device, the program code of a voice input step for inputting voice, and recognizing the input voice. Program code for the voice recognition step, program code for the determination step for determining which of the above communication means is to be used, recognition result obtained in the voice recognition step, and each connection destination abbreviation in the storage means And the program code of the verification step that selects the connection destination abbreviation according to the verification result. , Program code of the verification step, the determination in accordance with the determination result obtained in step, the program characterized by comprising a program code of the control step of limiting the matching candidates for each destination referred in the storage means. 23. Storage means describing the correspondence between a connection destination abbreviation and information specifying the connection destination, a communication means for making a call via a telephone line, and a facsimile communication for making a facsimile communication via a telephone line. A storage medium storing a program for controlling a communication device, comprising: a communication means for performing data communication via the Internet; and a program code of a voice input step for inputting voice, A program code of a voice recognition step for recognizing the voice, a program code of a determination step for determining which of the communication means is to be used, a recognition result obtained in the voice recognition step, and the storage means. Matching step abbreviations for each connection destination and selecting the connection destination abbreviation according to the matching result And a program code of the control step for limiting the verification candidate of each connection destination abbreviation in the storage means according to the determination result obtained in the determination step. A storage medium characterized by.