JP5046589B2 - Telephone system, call assistance method and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a system, method and program for enabling a hearing-impaired person or the like to correctly understand uttered contents on the telephone of communicating party and smoothly have a conversation on the telephone. <P>SOLUTION: The telephone system comprises a call control part 1 connected to a line for controlling the connection of a call, and a service center 4 connected to the call control part and provided with a voice recognition part 2 and an image preparation part 3. When the call from a caller is connected through the call connection part to the voice recognition part of the service center, the voice recognition part voice-recognizes the speech contents of a subscriber, converts the voice recognized result to character information, adds the reading information of the voice recognized result, and delivers it through the call control part to the image preparation part. The image preparation part prepares image data for which the voice recognized result and the reading information are combined and delivers them to the call control part. The call control part transmits them to the terminal of the subscriber on a call termination side, and the voice recognized result of the speech contents and the reading information are displayed on a screen at the terminal of the subscriber on the call termination side. <P>COPYRIGHT: (C)2008,JPO&amp;INPIT

Description

  The present invention relates to a telephone system, a call assistance method, and a program for recognizing a call content and displaying it on a terminal.

  This type of telephone system that recognizes the content of a call and displays it on a terminal is used, for example, in a telephone assistance system for a hearing impaired person. The typical basic configuration is that the content of a caller's utterance is recognized by a voice recognition device and displayed on a character display device. Conventional systems are roughly classified into two types. .

  One is that a dedicated terminal has a voice recognition function, recognizes the voice of the caller from the line, and displays it on a display device in the dedicated terminal (referred to as “terminal type”).

  The other is that a voice recognition device is provided on the center side, and voice and character information (text information) are transmitted to a terminal with a character display function and displayed on the terminal with a character information function (“center type”). Called).

  In Patent Document 1, when a caller's voice is received by a voice signal receiving means via a communication system and subtitle reception is selected by a voice / character conversion switch, the voice recognition device recognizes the received voice signal, Disclosed is a multimedia public telephone system that includes means for converting recognized speech into character data, displaying it as subtitles on the display, and conversely synthesizing the character string input by the caller as text and sending it to the other party. ing. In Patent Document 2, voice input from a microphone of a mobile phone is converted into digital voice data, supplied to a voice recognition unit of a computer for voice recognition, and a voice recognition result is returned to the mobile phone. After editing by the character editing unit of the telephone, it is displayed on the display, and when a correction command for the recognition result is input from the button operation unit of the mobile phone, the correction result is transmitted to the computer via a communication line such as the Internet. A character input device is disclosed that is returned and displayed on a display of a mobile phone and converted by a character editing unit of a computer (center) when conversion to kana-kanji is instructed, and the conversion result is returned and displayed on the display. Japanese Patent Application Laid-Open No. 2004-228561 includes a voice-recognized Japanese text display area and an input text display area as a voice typewriter that can easily specify a cursor when editing a voice-input sentence. Is disclosed.

Japanese Patent Laid-Open No. 10-224520 JP 2003-304331 A Japanese Examined Patent Publication No. 6-103457

  However, these conventional systems have the following problems in actual operation.

  A conversation freely spoken by an unspecified majority on the telephone is a process belonging to a class that is quite difficult to recognize even in speech recognition. In voice recognition on a telephone, the quality of voice passing through a network is significantly inferior to the quality of voice inputted through a microphone. This is because of deterioration in encoding, line echo, difficulty in noise removal, and the like due to transmission over a telephone line network.

  If the misrecognized speech recognition result is further converted into a kana-kanji mixed sentence, noise such as a conversion error is added, and it is difficult to estimate the original utterance content.

  Conventionally, although various voice recognition devices for a hearing impaired person and a telephone system including the device for a hearing-impaired person having a function of recognizing phone speech and displaying characters are proposed, Not reached. This is because, generally speaking, it is difficult to sequentially recognize speech content of a telephone call and performance degradation due to erroneous recognition is expected.

  For hearing-impaired persons, listening to the content of telephone utterances is an urgent problem, and an early solution and practical application is desired.

  Therefore, an object of the present invention is to provide a system, a method, and a program that enable a hearing impaired person or the like to accurately understand the utterance contents of the other party's telephone, thereby enabling a smooth call. is there.

  In order to solve the above-described problems, the invention disclosed in the present application is generally configured as follows.

  A system according to one aspect of the present invention includes a unit that recognizes a voice signal input from a first terminal, a unit that generates reading information of the voice recognition result, and at least the reading information. Means for displaying on a second terminal which is a call partner of the first terminal.

  The system according to the present invention includes means for creating screen data including character information of a speech recognition result and the reading information, and the image data is transmitted to the second terminal.

  In the system according to the present invention, the voice recognition unit, the reading information generation unit, and the screen data generation unit are provided in a service center connected to a line via a call connection unit.

  The system according to the present invention further includes a voice recognition unit including the voice recognition unit and the reading information generation unit, and the voice recognition unit receives the phone voice from the first terminal and recognizes the voice. A speech recognition result is converted into character information (text), reading information of the speech recognition result is further generated, and an image data creation unit for creating the recognition result converted into text and screen data including the reading information is provided. The screen data is transmitted to the second terminal, and a screen including the recognition result converted into text and the reading information is displayed on the second terminal.

  In the system according to the present invention, the second terminal may include a unit that generates reading information of the voice recognition result.

  A system according to the present invention includes a call control unit that is connected to a line and performs call connection control, and a service center that is connected to the call control unit and includes a voice recognition unit and a screen creation unit. When a call from a caller is connected to the voice recognition unit of the service center via the call control unit, the voice recognition unit recognizes the content of the call from the caller and performs a voice recognition result. Is converted into character information, and reading information of a speech recognition result is added and passed to the screen creation unit via the call control unit. The screen creation unit is a screen data combining the speech recognition result and the reading information. Is generated and passed to the call control unit, and the call control unit transmits the call to the subscriber's terminal. Is displayed on the screen.

  In the system according to the present invention, the service center includes a reading generation unit separately from the voice recognition unit, and after the call is connected, the voice recognition unit of the service center includes a call content from a caller. The speech recognition result is transferred to the call control unit, the call control unit sends the speech recognition result to the reading generation unit of the service center, and the reading generation unit Reading is estimated from the result, reading information is generated, and the reading information is transmitted to the call control unit. The call control unit includes character information of a recognition result output from the voice recognition unit of the service center; , The reading information from the reading generation unit of the service center is sent to the screen creation unit of the service center, and the screen creation unit obtains screen data combining the voice recognition result and the reading information. It is formed.

  In the system according to the present invention, a reading generation unit is provided in the terminal on the called side separately from the voice recognition unit, the voice recognition unit of the service center does not generate reading information, and the screen generation unit The screen data of the voice recognition result may be generated, and the reading generation unit of the terminal on the called side may generate reading information by estimating the reading from the voice recognition result.

  In the present invention, the voice from the called terminal is transmitted to the calling terminal, and is received and reproduced by the calling terminal.

  A service center according to another aspect of the present invention is a service center including a voice recognition unit connected to a call control unit and a screen creation unit, and a call from a caller passes through the call control unit. When connected to the voice recognition unit, the voice recognition unit recognizes the content of the call from the caller, converts the voice recognition result into character information, and adds reading information of the voice recognition result. The screen creation unit passes the call control unit to the screen creation unit, and the screen creation unit creates screen data that combines the speech recognition result and the reading information, and passes it to the call control unit. To the terminal of the called subscriber.

  The service center according to the present invention further includes a reading generation unit, and after the call is connected, the voice recognition unit recognizes the content of the call from the caller and uses the voice recognition result as the call control unit. The reading generation unit estimates the reading from the speech recognition result, transmits the reading information to the call control unit, and the call control unit receives the voice recognition of the service center. The output result from the section and the reading information from the reading generation section are sent to the screen creation section of the service center, and the screen creation section creates screen data that combines the voice recognition result and the reading information. It may be.

  In the present invention, the character information is obtained by converting the voice recognition result to Kana-Kanji, and the reading information is obtained by estimating the reading from the voice recognition result, and includes at least one of Hiragana, Roman characters, and phonetic symbols. It is written in one.

According to yet another aspect of the present invention,
Recognizing a voice signal input from the first terminal;
Generating reading information of the speech recognition result;
Displaying at least the reading information on a second terminal that is a call partner of the first terminal;
including.

  In the method according to the present invention, the image data is transmitted to the second terminal, including the step of creating screen data including character information of a speech recognition result and the reading information.

  In the method according to the present invention, the second terminal may generate reading information of the voice recognition result.

The method according to the present invention includes a call control unit that is connected to a line and performs call connection control, and a service center that is connected to the call control unit and includes a voice recognition unit and a screen creation unit. The call recognition method of claim 1, wherein when a call from a caller is connected to the voice recognition unit of the service center via the call control unit, the voice recognition unit Recognizes the content, converts the speech recognition result into character information, adds reading information of the speech recognition result, passes to the screen creation unit via the call control unit,
The screen creation unit creates screen data combining the voice recognition result and the reading information and passes it to the call control unit, and the call control unit transmits to the terminal of the called subscriber,
At the called subscriber's terminal, the voice recognition result and the reading information of the call contents are displayed on the screen.

  In the method according to the present invention, the service center includes a reading generation unit separately from the voice recognition unit, and after a call is connected, the voice recognition unit of the service center receives a call from a caller. Recognizes the content and passes the speech recognition result to the call control unit, the call control unit sends the speech recognition result to the reading generation unit of the service center, and the reading generation unit Reading is estimated, and the reading information is transmitted to the call control unit. The call control unit receives the character information of the recognition result output from the voice recognition unit of the service center and the reading information from the reading generation unit. May be sent to the screen creation unit of the service center, and the screen creation unit may create screen data that combines the recognition result and the reading information.

The program according to the present invention is:
Processing for recognizing a voice signal input from the first terminal;
Processing to generate reading information of speech recognition results;
It comprises a program for causing a computer to execute at least the process of displaying the reading information on a second terminal that is a call partner of the first terminal.

  The program according to the present invention includes a program for causing the computer to execute processing for creating character data of a speech recognition result and screen data including the reading information.

A program according to the present invention includes a voice recognition unit connected to a call control unit and a screen creation unit, and when a call from a caller is connected to the voice recognition unit via the call control unit. The voice recognition unit recognizes the content of the call from the caller, converts the voice recognition result into character information, and adds reading information of the voice recognition result to the screen creation unit via the call control unit. To
The screen creation unit creates screen data that combines the voice recognition result and its reading information, passes it to the call control unit, and transmits it to the terminal of the called subscriber via the call control unit. The program comprises a program that causes a computer constituting a service center to execute the processing of the voice recognition unit and the screen creation unit.

  In the program according to the present invention, the service center includes a reading generation unit separately from the voice recognition unit, and after the call is connected, the voice recognition unit recognizes the content of the call from the caller. The voice recognition result is converted into character information and sent to the reading generation unit via the call control unit. The reading generation unit estimates a reading from the voice recognition result, and reads the reading information into the call control unit. To the computer that constitutes the service center to be transmitted to the computer, and includes a program for executing the processing of the voice recognition unit, the screen creation unit, and the reading generation unit.

  According to the present invention, a hearing-impaired person (caller) can smoothly talk with a healthy person (caller) by telephone. The reason for this is that in the present invention, even if the speech recognition apparatus gives a wrong recognition result, the hearing impaired person (calling party) infers the error tendency from the reading and correctly utters. This is because the contents can be understood.

  The best mode for carrying out the present invention will be described below. In recent years, telephone terminals with a screen display function that allow browsing of not only character information (text information) but also www (world wide web), such as an L-mode and an IP phone using an IP network, have been used. If such a terminal is used, a hearing-impaired person can use a so-called telephone line by e-mail or the like. However, if the content of the caller's utterance is displayed on the screen, the hearing person who has the ability to speak A person with a disability is thought to be able to use the phone in the original sense. However, as described above, it is difficult to recognize conversational speech by telephone speech recognition, and it is expected that considerable misrecognition and erroneous conversion may occur.

  Therefore, the present invention provides a speech recognition unit (2) that recognizes speech content from a calling terminal (5) and converts it into character information and reading information, and a screen for a terminal with a screen display. A screen creation unit (3) to be created and a call control unit (1) for managing a call and controlling the voice recognition unit (2) and the screen creation unit (3) are created by the screen creation unit (3). The screen data includes not only the utterance content (voice recognition result) of the speaker but also reading information, which is transmitted to the called terminal (6) and displayed. The reading information is arbitrary as long as it represents the reading of the character information of the speech recognition result of the utterance content, such as hiragana, kana, romaji, and phonetic symbols.

  When the present invention is applied to a telephone assistance system for a hearing impaired person, in a telephone assistance system for a hearing impaired person comprising a telephone terminal, a public network, a voice recognition unit, and a telephone terminal capable of displaying characters and images, speech speech is transmitted. Voice recognition is performed and converted into character information to which a voice recognition result and voice reading information are added and displayed on a telephone terminal with a display function for a hearing impaired person, thereby enabling conversation by the telephone of the hearing impaired person.

  The outline of the operation of the present invention will be explained. When a call from the terminal (5) on the calling side is connected to the voice recognition unit (2) of the service center (4) via the call control unit (1). The voice recognition unit (2) recognizes the content of the subscriber's call and converts it into character information. At this time, since there is a possibility of erroneous recognition, the reading information is added to the screen together with the recognition result. Delivered to the creation unit (3). The screen creation unit (3) processes and creates a screen combining the voice recognition result and the reading information, and transmits it to the called subscriber terminal. On the terminal (6) of the service subscriber on the called side, the received screen (voice recognition result and reading information of the call content) is displayed. For this reason, even if the recognition and conversion in voice recognition are incorrect, it is associated with the reading information, so that it is possible to facilitate the decoding of the original call contents and to help the decoding effectively. In addition, it is good also as a structure provided with the reading production | generation part which produces | generates reading information separately from a speech recognition part (2). Alternatively, reading information may be generated by estimating the reading from the voice recognition result at the terminal (6) of the called subscriber. Examples will be described below.

  FIG. 1 is a diagram showing the configuration of an embodiment of the present invention. Referring to FIG. 1, the telephone assistance system of the present embodiment includes a service center 4 including a call control unit 1, a voice recognition unit 2, and a screen creation unit 3. The processing of the voice recognition unit 2 and the screen creation unit 3 may be realized by a program executed on a computer.

  FIG. 2 is a diagram illustrating an example of the configuration of the call control unit 1. Referring to FIG. 2, the call control unit 1 includes a control unit 10, a voice storage unit 11, and a media conversion unit 12, manages a call, and transmits received voice to the voice recognition unit 2.

  The control unit 10 performs overall control of the call control unit 1, manages incoming calls, and transmits and receives voices.

  The voice storage unit 11 has a function of storing voices on the calling side and the called side, and also stores guidance message voices for the caller side.

  The media conversion unit 12 has a protocol conversion function and performs voice data conversion on both the called side and the calling side.

  The control unit 10 manages the call, temporarily stores the received voice in the voice storage unit 11, and transmits the voice to the voice recognition unit 2.

  Referring to FIG. 1 again, the speech recognition unit 2 recognizes the received speech and outputs the output result (speech recognition result and reading information) via the control unit 10 (see FIG. 2) of the call control unit 1. To the screen creation unit 3. In the present embodiment, the voice recognition unit 2 performs voice recognition by any known method for recognizing the voice of an unspecified speaker, for example, for a recognition result in which a word is determined using a word dictionary. A kana-kanji conversion is performed to generate a sentence (text sentence) corresponding to the utterance content, and further, a process of adding reading information to the character information of the recognition result is executed.

  In the screen creation unit 3, the received output result (speech recognition result and reading information) of the speech recognition unit 2 can be displayed on a telephone terminal with a screen display function (mainly page description such as HTML (HyperText Markup Language) language). Language).

  When the control unit 10 (see FIG. 2) of the call control unit 1 receives the data (screen display data) output from the screen creation unit 3, it transmits the screen display data to the called line.

  When the telephone terminal 6 on the called side receives the screen display data, the voice recognition result uttered on the speaking side and the reading information thereof are displayed on the screen. For this reason, even if there are some errors in the speech recognition result, it can be easily understood.

  The caller's voice is received by the call control unit 1, temporarily stored in the voice storage unit 11 (see FIG. 2) in the call control unit 1, and then in a form that matches the caller side protocol. Sent to the side.

  In this way, a conversation can be made between the calling side of the healthy person and the calling side of the deaf person.

  FIG. 3 is a flowchart for explaining the operation of this embodiment. The operation of this embodiment will be described with reference to FIG. 1, FIG. 2, and FIG. However, the call is already connected.

  The call control unit 1 receives a voice signal from the calling terminal 5 (step S1).

  The input voice signal is registered and managed by the control unit 10 of the call control unit 1, and the voice is stored in the voice storage unit 11 of the call control unit 1 (step S2).

  The control unit 10 of the call control unit 1 requests the voice recognition unit 2 of the service center 4 for voice recognition processing and transmits the accumulated voice (step S3).

  In the voice recognition unit 2 of the service center 4, the voice signal received from the control unit 10 of the call control unit 1 is voice-analyzed by acoustic analysis or the like, and the output result is output as output data including the recognition result and reading information. To the control unit 10 of the unit 1.

  The control unit 10 of the call control unit 1 transmits the output result received from the voice recognition unit 2 of the service center 4 to the screen creation unit 3 of the service center 4 (step S4).

  The screen creation unit 3 of the service center 4 creates screen information data composed of the voice recognition result (mainly kana-kanji mixed text) and the reading information based on the received output data, and the control unit of the call control unit 1 10 (step S5).

  Upon receiving the created screen information, the control unit 10 of the call control unit 1 transmits it to the incoming call line (step S6).

  The screen display data is received by the called terminal (phone terminal with screen display function) 6, and the recognition result and the reading information are displayed on the screen of the called terminal 6 (step S7).

  In this way, when the incoming call side telephone terminal 6 with screen display function receives and displays the screen information, the voice recognition result (for example, kana kanji text) and the reading information (for example, hiragana notation) are displayed. If the speech recognition result is correct, there is no problem at all, but even if the recognition result of kana / kanji sentences contains errors, the correct correct utterance content can be estimated from the reading information by observing the original correct utterance content. You can turn it on.

  Referring to FIG. 3B, the recipient's voice passes through the line and reaches the call control unit 1 (step S11).

  The received voice is stored in the voice storage unit 11 of the call control unit 1 by the control unit 10 of the call control unit 1 (step S12).

  Based on the information registered in step S2 in FIG. 3A, the media conversion unit 12 of the call control unit 1 transmits to an appropriate caller in an appropriate procedure according to the terminal 5 on the call side ( Step S13).

  The calling side terminal 5 receives and reproduces the called side uttered voice (step S14).

FIG. 4 is a diagram showing an example of the screen of the called terminal (telephone terminal with a screen display function) 6 of one embodiment of the present invention. For example, from the caller
“Can i-mode (registered trademark) be used?”
Then, it is assumed that there is an input to the call control unit 1 (step S1 in FIG. 3A).

  In the call control unit 1, the voice signal and call information are registered in the control unit 10 of the call control unit 1, and the voice signal is stored in the voice storage unit 11 of the call control unit 1 (step S2).

  The control unit 10 of the call control unit 1 requests the voice recognition unit 2 of the service center 4 for recognition processing and transmits the accumulated voice signal (step S3).

When the voice recognition unit 2 of the service center 4 receives the voice signal from the voice storage unit 11, it analyzes the voice signal,
Recognition result: “Is it not possible to use love alone?” And
Reading information: "Is it impossible to use Aiyodo"
Is output to the control unit 10 of the call control unit 1. Although not particularly limited, in this example, the recognition result is a sentence converted to Kana-Kanji, and the reading information is in Hiragana notation.

  The control unit 10 of the call control unit 1 requests the screen creation unit 3 of the service center 4 for screen information creation processing and transmits the output data (step S4).

  Based on the received data, the screen creation unit 3 of the service center 4 creates screen information composed of character information and reading information including the voice recognition result, and transmits the screen information to the control unit 10 of the call control unit 1 (step S5). ).

  Upon receipt of the created screen information, the control unit 10 of the call control unit 1 transmits it to the called-side line (step S6).

  The screen information is received by the telephone terminal 6 with the screen display function on the called side, and the utterance content recognition result and the utterance content reading information are displayed (step S7).

  In FIG. 4, the utterance content recognition result is displayed as “love and self-existence”, and if the caller does not know how to read “self-existence (udo)”, the callee determines what “aimadokatsu” means You will be troubled.

  However, if you look at the reading information at the bottom, the reading information displays “Audou”, so even if it contains errors in speech recognition (for example, word determination processing or kana-kanji conversion in speech recognition) Even when there is an error in the processing, the correct utterance content “i-mode” can be determined by observing the correct utterance content from the reading information.

  Next, the incoming call side (for example, a deaf person) speaks (step S11 in FIG. 4B).

  The control unit 10 of the call control unit 1 stores the received voice in the voice storage unit 11 (step S12).

  The media conversion unit 12 of the call control unit 1 transmits the accumulated voice based on the registration information to the caller (step S13). At this time, for example, if the called party is HTTP (Hyper Text Transport Protocol) and the calling party is VoIP, it is converted to RTP (Real-time Transport Protocol), and if the calling party is PSTN (Public Switched Telephone Networks) To the digital hardware line.

  The caller's voice is reproduced at the caller side terminal 5 (step S14). In this way, it is possible to have a conversation between the calling party (healthy person) and the deaf person (calling side).

  Next, a second embodiment of the present invention will be described. FIG. 5 is a diagram showing the configuration of the second exemplary embodiment of the present invention. In this embodiment, the reading information is added by the screen creation unit 31 on the service center 4 'side.

  In this case, it is not necessary for the speech recognition unit 2 to output a reading, so that the speech recognition unit 2 can use the existing one as it is.

  Referring to FIG. 5, this embodiment includes a service center 4 ′ having a call control unit 1, a voice recognition unit 2, a screen creation unit 31, and a reading generation unit 32.

  The call control unit 1 includes a control unit 10, a voice storage unit 11, and a media conversion unit 12, similarly to the configuration illustrated in FIG. 2. However, the control unit 10 of the call control unit 1 is connected to the voice recognition unit 2, the screen creation unit 31, and the reading generation unit 32.

  After the call is connected, the voice recognition unit 2 recognizes the received voice and transmits the output result (only the voice recognition result) to the control unit 10 of the call control unit 1.

  The control unit 10 of the call control unit 1 holds the output result in the voice storage unit 11 and sends it to the reading generation unit 32 of the service center 4 ′.

  The reading generation unit 32 of the service center 4 ′ estimates the reading from the voice recognition result and transmits the reading information to the control unit 10 of the call control unit 1.

  The control unit 10 of the call control unit 1 sends the output result from the voice recognition unit 2 of the service center 4 ′ and the reading notation from the reading generation unit 32 to the screen creation unit 31 of the service center 4 ′.

  The screen creation unit 31 of the service center 4 ′ processes the voice recognition result and the reading information into a form (mainly a page description language such as HTML language) that can be displayed on the telephone terminal 6 with a screen display function, and transmits it to the control unit 10. To do.

  When the control unit 10 of the call control unit 1 receives the data output from the screen creation unit 31 of the service center 4 ′, it transmits it to the called line.

  Note that the caller's voice is transmitted to the caller as it is as in the above embodiment.

  FIG. 6 is a flowchart for explaining the operation of the second embodiment of the present invention. The operation of the second embodiment of the present invention will be described with reference to FIG. 5, FIG. 2, and FIG.

  The call control unit 1 receives a voice signal from the caller (step S21).

  The input voice signal is registered and managed by the control unit 10 of the call control unit 1, and the voice is stored in the voice storage unit 11 of the call control unit 1 (step S22).

  The control unit 10 of the call control unit 1 transmits the accumulated voice to the voice recognition unit 2 of the service center 4 '(step S23).

  The voice recognition unit 2 of the service center 4 ′ analyzes and recognizes the voice signal received from the control unit 10 of the call control unit 1, and outputs the output result to the control unit 10 of the call control unit 1 using the recognition result as output data. .

  The control unit 10 of the call control unit 1 transmits the received output result to the reading generation unit 32 of the service center 4 '(step S24).

  The reading generation unit 32 of the service center 4 'estimates reading information from the received data, and transmits the result to the control unit 10 of the call control unit 1 (step S25).

  The control unit 10 of the call control unit 1 transmits the output result of the voice recognition unit 2 of the service center 4 ′ and the output result of the reading generation unit 32 to the screen creation unit 31 of the service center 4 ′ (step S26). .

  The screen creation unit 31 of the service center 4 ′ creates screen information data composed of character information and reading information containing the voice recognition result based on the received data, and transmits it to the control unit 10 of the call control unit 1 ( Step S27).

  Upon receiving the created screen information, the control unit 10 of the call control unit 1 transmits it to the incoming call line (step S28).

  The screen display data is received by the terminal 6 (the telephone terminal with the screen display function on the called side), and the utterance content recognition result and the utterance content reading information are displayed on the telephone terminal 6 with the screen display function on the called side (step S29). .

  When the incoming call side telephone terminal 6 with screen display function receives and displays the screen information, the voice recognition result and the reading information are displayed on the same screen. Even if it is included, the correct pronunciation can be observed from the reading information and the correct utterance content can be determined.

  Next, the telephone voice of the called party passes through the line and reaches the call control unit 1 (step S11 in FIG. 3B).

  The received voice is stored in the voice storage unit 11 by the control unit 10 of the call control unit 1 (step S12 in FIG. 3B).

  Based on the information registered in step S1, the media conversion unit 12 of the call control unit 1 transmits to an appropriate caller in an appropriate procedure according to the caller (step S13 in FIG. 3B).

  The calling side terminal 5 receives and reproduces the voice uttered by the called side (step S14 in FIG. 3B).

  As a third embodiment of the present invention, the reading information may be generated on the terminal 6 side of the called side. In this case, it is not necessary to generate reading information in the voice recognition unit 2 or the reading generation unit 32 of the service centers 4 and 4 ′ in FIGS. 1 and 5, and it is not necessary to add reading information in the screen creation unit 3. This reduces the processing load and burden on the service center side. Regarding the processing procedure of the third embodiment of the present invention, the reading center is not handled on the service center side, and the reading generation unit is activated on the terminal 6 side that has received the screen information on which the recognition result is displayed. Generate information. Other processing follows the procedure of the above embodiment.

  The present invention is suitable for providing telephone services such as welfare and social services.

  Although the present invention has been described with reference to the above-described embodiments, the present invention is not limited to the configurations of the above-described embodiments, and various modifications that can be made by those skilled in the art within the scope of the present invention. Of course, including modifications.

It is a figure which shows the structure of one Example of this invention. It is a figure which shows the structure of the call control part of one Example of this invention. It is a flowchart for demonstrating operation | movement of one Example of this invention. It is a figure which shows the example of a screen display of the telephone terminal with a screen display function of the callee side of one Example of this invention. It is a figure which shows the structure of another Example of this invention. It is a flowchart for demonstrating operation | movement of another Example of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Call control part 2 Voice recognition part 3, 31 Screen creation part 32 Reading production | generation part 4, 4 'Service center 5 Calling side terminal 6 Calling side terminal 10 Control part 11 Voice storage part 12 Media conversion part

Claims (14)

A call control unit that is connected to the line and controls call connection;
A service center connected to the call control unit, comprising a voice recognition unit and a screen creation unit;
With
Call from caller, via the call control unit, when connected to the speech recognition unit of the service center, the voice recognition unit, the utterance content from the calling party recognizes speech, The recognition result in which the word is determined by the voice recognition is converted into character information corresponding to the utterance content, and the voice recognition unit adds reading information to the character information, and the call control unit To the screen creation unit via
The screen creation unit creates screen data that combines the character information corresponding to the utterance content and its reading information and passes it to the call control unit,
The call control unit transmits to the called terminal,
Telephone system the called terminals, wherein you screen and reading the character information data corresponding to the utterance contents, characterized in that. The service center includes a reading generation unit separately from the voice recognition unit,
After the call from the caller is connected, in the voice recognition unit of the service center, the utterance content from the caller's voice recognition, the recognition result of the word has been determined by the voice recognition to the The voice recognition unit does not generate reading information for the character information corresponding to the utterance content, and the character information corresponding to the utterance content is sent to the call control unit. Hand over,
The call control unit sends character information corresponding to the utterance content to the reading generation unit of the service center,
The reading generation unit generates reading information by estimating reading from character information corresponding to the utterance content, and transmits the reading information to the call control unit,
The call control unit generates character information corresponding to the utterance content output from the voice recognition unit of the service center and reading information from the reading generation unit of the service center, and creates the screen of the service center. To the department,
The screen creating section, the telephone system of claim 1, wherein the creating the character information corresponding to the speech content and a screen data combined thus read information, it is characterized. Separately from the voice recognition unit, a reading generation unit is provided in the terminal on the called side,
In the voice recognition portion of the service center, it does not generate read information, the screen creation unit creates the screen data of the character information corresponding to the speech content from the speech recognition unit,
Telephone system according to claim 1, wherein the generating unit attached reading of the called terminal generates the reading by estimating information read from the character information corresponding to the uttered contents, it is characterized. Speech from the called terminal is sent to the calling terminal, the telephone system according to claim 1, wherein the received at the calling side terminal is reproduced, characterized in that. A service center including a voice recognition unit connected to the call control unit and a screen creation unit,
Words a call from the caller is connected to the speech recognition unit via the call control unit, the voice recognition unit, the utterance content from the calling party voice recognition by the voice recognition Is converted into character information corresponding to the utterance content , and the voice recognition unit adds reading information to the character information, and the screen is transmitted via the call control unit. To the creation department,
The screen creation unit creates screen data that combines the character information corresponding to the utterance content and its reading information, and transmits the screen data to the called terminal via the call control unit. A featured service center. 6. The service center according to claim 5 , further comprising a reading generation unit separately from the voice recognition unit,
After the call from the caller is connected, the voice recognition unit, the utterance content of the caller's voice recognition, to the speech content with respect to the recognition result word has been determined by the voice recognition The voice recognition unit does not generate reading information for the character information corresponding to the utterance content, and reads the character information corresponding to the utterance content via the call control unit. To the generator,
The reading generation unit generates reading information by estimating reading from character information corresponding to the utterance content, and transmits the reading information to the call control unit,
The call control unit sends the character information corresponding to the utterance content output from the voice recognition unit of the service center and the reading information from the reading generation unit to the screen creation unit of the service center. A service center characterized by this. The character information is a kana-kanji converted recognition result in which a word is determined by the voice recognition,
The telephone according to any one of claims 1 to 3 , wherein the reading information represents character information corresponding to the utterance content in at least one of hiragana, romaji, and phonetic symbols. system. A call control unit that is connected to the line and controls call connection;
A service center connected to the call control unit and comprising a voice recognition unit and a screen creation unit;
A call assistance method for a telephone system comprising:
Call from caller, via the call control unit, when connected to the speech recognition unit of the service center, the voice recognition unit, the utterance content from the calling party recognizes speech, The recognition result in which the word is determined by the voice recognition is converted into character information corresponding to the utterance content , and the voice recognition unit adds reading information to the character information, and the call control unit To the screen creation unit via
The screen creation unit creates screen data combining the character information corresponding to the utterance content and its reading information and passes it to the call control unit, and the call control unit transmits to the terminal on the called side,
The called terminal, the you screen displaying information read the character information corresponding to the uttered contents, call the auxiliary method for a telephone system, characterized in that. The service center includes a reading generation unit separately from the voice recognition unit,
After the call from the caller is connected, the voice recognition unit of the service center, the utterance content from the caller's voice recognition, the recognition result of the word has been determined by the voice recognition to the The voice recognition unit does not generate reading information for the character information corresponding to the utterance content, and the character information corresponding to the utterance content is sent to the call control unit. Hand over,
The call control unit sends character information corresponding to the utterance content to the reading generation unit of the service center, and the reading generation unit estimates a reading from the character information corresponding to the utterance content, and reads the reading information. and transmitted to the call control unit, the call control unit, and the character information corresponding to the speech content output from the speech recognition unit of the service center, the read information from the read put generating unit, the service 9. The telephone system call assistance according to claim 8 , wherein the screen creation unit creates screen data that combines character information corresponding to the utterance content and its reading information. Method. 9. The telephone system call assisting method according to claim 8, wherein the voice from the called terminal is transmitted to the calling terminal, and is received and played back by the calling terminal. The character information is a kana-kanji converted recognition result in which a word is determined by the voice recognition,
10. The telephone system call assistance according to claim 8 or 9 , wherein the reading information represents character information corresponding to the utterance content in at least one of hiragana, romaji, and phonetic symbols. Method. A voice recognition unit connected to the call control unit, and a screen creation unit;
When a call from a caller is connected to the voice recognition unit via the call control unit,
The voice recognition unit, said utterance content from the caller voice recognition, converted into character information corresponding to the uttered contents against recognition result word is determined by the speech recognition, yet the voice recognition parts adds information read to the text information, executes the pass to process the viewing preparation unit through the call control unit,
The screen creation unit creates screen data combining character information corresponding to the utterance content and reading information, passes the screen data to the call control unit, and transmits the call data to the called terminal via the call control unit. the computer constituting the service center to execute the processing program for executing the respective processing of the viewing preparation unit and the voice recognition unit. In claim 1 wherein the program,
The service center includes a reading generation unit separately from the voice recognition unit,
After the call from the caller is connected, the voice recognition unit, the utterance content from the calling party speech recognition, the speech content for the recognition result word has been determined by the voice recognition The speech recognition unit does not generate reading information for the character information corresponding to the utterance content, and the character information corresponding to the utterance content is read via the call control unit. run the feeding that process to put generation unit,
The read put generating unit estimates read from the character information corresponding to the speech content, the read information, the computer constituting the service center to execute the process of transmitting to the call control unit, and the speech recognition unit program for executing the processes of generating portion attached to read the said screen creating unit. The character information is a kana-kanji converted recognition result in which a word is determined by the voice recognition,
The read information is, Hiragana those estimated reading character information corresponding to the uttered contents, Romaji, is obtained by representation of at least one of phonetic symbols, according to claim 1 2 or 13, characterized in that The program described in.

Images