JP6342972B2 - Communication system and communication method thereof - Google Patents

Description

  The present invention relates to a communication system and a communication method thereof, and more particularly, to a communication system and a communication method thereof suitable for use in receiving speech from a telephone terminal and performing speech translation.

  In recent years, communication infrastructure such as the Internet has been developed on a global scale, and the globalization of communication is rapidly progressing. Under such circumstances, automatic speech translation is attracting attention as information processing technology advances.

  Automatic speech translation recognizes the voice of the input person, converts it into text, automatically translates the text in the input language into text in the target language, and converts the translated text into speech in the target language And a system that outputs.

  As a technique using such automatic speech translation, for example, there is Patent Document 1. In Patent Document 1, the urgency level and language type are determined from voice data received from a mobile terminal. When the urgency level is high, an interpreter who is familiar with various languages is selected. When the urgency level is low, machine translation is performed. A technique for causing a server to perform machine translation is disclosed.

Japanese Patent Laid-Open No. 2006-66983

  When speech translation is performed by automatic speech translation, it is necessary to specify an input language (source language) and an output language (translation language) in advance. When using a mobile terminal such as a smartphone as in Patent Document 1, a language can be specified by tapping the screen. However, in the case of a general fixed telephone, there is a problem that it cannot be specified in the same way.

  Further, in face-to-face speech interpretation, it is conceivable to use an automatic speech translation system to listen to speech translated from the other party's utterance while alternately passing one telephone. In such a case, there is a problem that the beginning of the output voice may not be heard due to variations in the delivery timing of the telephone.

  Further, regarding the specification of the language type, in the interpreting service system described in Patent Document 1, a keyword such as “English” or “French” is prepared for each language type, and a security guard or a foreigner speaks the keyword. Thus, the foreign language is specified (paragraph number 0059).

  However, the determination of the language type described in Patent Document 1 is a detour, a load is imposed on the processing system, and it takes extra time for communication, so that one telephone is alternately transferred in the face-to-face speech interpretation. In application, there is a problem that it cannot be delivered smoothly.

  The present invention was made to solve the above problems, and its purpose is to specify a language type with a simple specification when a telephone and an automatic speech translation system exchange voice over a telephone line. In a face-to-face voice interpreter, when a single telephone is handed over alternately, a communication system that allows smooth handing over, allows the handed-over handset to hear the appropriate voice, and prevents missed voices, and It is to provide the communication method.

  A communication system according to the present invention is a communication system in which a telephone terminal and a communication server are connected by a telephone line, and the telephone terminal transmits a call to and from the communication server and inputs from a key device. Tone Multi-Frequency) signal is generated and transmitted to the communication server. The communication server is transmitted by the language processing unit that performs speech translation from the first language to the second language, and transmitted by the telephone line. Means for translating the first language voice related to the incoming voice signal into the second language and transmitting it to the telephone terminal, voice data of the speaker language, and the voice of the translated voice obtained by translating the voice data A call data table for storing data. When the communication server receives the first DTMF signal from the telephone terminal, the communication server receives the first DTMF signal until it receives the second DTMF signal operated and transmitted by another speaker. The voice related to the voice signal received is stored in the call data table as voice data in the language represented by the first DTMF signal, and the language represented by the first DTMF signal is transmitted after the second DTMF signal is transmitted. The speech signal related to the speech data of the translated speech obtained by translating the speech data into the language represented by the second DTMF signal is transmitted to the telephone terminal.

  According to the present invention, when a telephone and an automatic speech translation system exchange voice over a telephone line, the language type can be designated with simple designation, and one telephone is alternately delivered in face-to-face speech interpretation. In addition, it is possible to provide a communication system and a communication method capable of smoothly delivering, allowing a person who has received the handset to hear the sound with an appropriate voice, and preventing the voice from being missed.

1 is an overall configuration diagram of a communication system. It is a functional block diagram of a telephone terminal. It is a block diagram of a communication server. It is a figure explaining the state which a communication control part recognizes. It is a figure explaining the event which a communication control part receives. It is a state transition diagram showing the state transition due to an event. It is a figure which shows the process matrix by a state and the event which generate | occur | produces. It is a figure explaining the specification of the button corresponding | compatible table. It is a figure explaining the specification of the call state table. It is a figure explaining the specification of the call data table. FIG. 2 is a schematic sequence diagram showing delivery between each component of a communication system and a state of the system (part 1). It is a general | schematic sequence diagram which shows the delivery between each component of a communication system, and the state of a system (the 2). It is a general | schematic flowchart which shows the process of a communication server. It is a flowchart which shows an audio | voice analysis process. It is a flowchart which shows a repeat process.

  Embodiments according to the present invention will be described below with reference to FIGS.

First, the configuration of a communication system according to an embodiment of the present invention will be described with reference to FIGS. 1 and 3.
FIG. 1 is an overall configuration diagram of a communication system.
FIG. 2 is a functional configuration diagram of the telephone terminal.
FIG. 3 is a configuration diagram of the communication server.

  As shown in FIG. 1, the communication system of the present embodiment is a form in which a telephone terminal 10 and a communication server 100 are connected by a telephone line 5.

  The telephone line 5 may be a public line or a private line that is switched by a private branch exchange (PBX) in a company. Also, an analog line or a digital line may be used.

  The telephone terminal 10 needs to be a telephone capable of transmitting a DTMF (Dual-Tone Multi-Frequency) signal (so-called push signal). A DTMF signal is a composite signal sound of two voice frequency bands of low and high groups, consisting of numbers from 0 to 9 and symbols of *, #, A, B, C, and D in total. A signal to be transmitted. The line connected to the telephone terminal 10 is a line that can transmit a DTMF signal.

  As shown in FIG. 2, the telephone terminal 10 includes a receiver 11, a transmitter 12, a key device 13, an audio signal converter 14, a connection controller 15, and a DTM signal generator 16. The receiver 11 and the transmitter 12 are a speaker for receiving and a microphone for transmitting, respectively. The voice signal conversion unit 14 is a part that converts an electrical signal from the telephone line into voice and converts the voice from the transmitter 12 into an electrical signal. The connection control unit 15 is a part that controls incoming / outgoing calls to connect / disconnect the line to / from the telephone terminal 10. The DTM signal generator 16 is a device that generates a DTM signal having a predetermined frequency in accordance with an input from the key device 13.

  A person who uses the telephone terminal 10 can input his / her language type from the key device 13 according to a button of a rule defined for each language. For example, [#] [1] is Japanese, [#] [2] is English, and [#] [3] is Chinese. The telephone terminal 10 converts it into a DTM signal and transmits it to the communication server 100 via the telephone line 5 as a signal for determining the language of the user of the telephone terminal 10.

  The communication server 100 decodes a voice signal transmitted via the telephone line 5, converts it into voice, and translates it into voice according to a designated language (hereinafter referred to as “translation language”) (hereinafter referred to as “translation language”). This is a device that encodes the translated speech and transmits it to the telephone terminal 10 via the telephone line 5.

  As illustrated in FIG. 3, the communication server 100 includes a reception control unit 110, a transmission control unit 120, a communication control unit 150, a database access unit 160, a language processing unit 200, and a database 300.

  As shown in FIG. 3, the reception control unit 110 includes a decoder 111, a control signal analysis unit 112, a DTMF signal analysis unit 113, a voice analysis unit 114, and a call data output unit 115. The reception voice signal and DTMF signal are received. This is the part that determines and decodes and writes the received information to the database. The reception control unit 110 receives a control signal, a voice signal, or a DTMF signal from the telephone terminal 10 via a telephone line and decodes it by the decoder 111. Then, it is discriminated whether it is a control signal, a DTMF signal, or a voice, and if it is a control signal, it recognizes the call of the call and reports it to the call data output unit 115. In the case of a DTMF signal, the DTMF signal analysis unit 113 analyzes the frequency included in the DTMF signal, recognizes which button is pressed on the telephone terminal 10, and reports it to the call data output unit 115. In the case of voice, the voice analysis unit 114 determines one sentence and reports the voice data to the call data output unit 115.

  The call data output unit 115 generates a new call ID when a new call is called. Further, referring to the button correspondence table 310, the language ID of the speaker corresponding to the pressed button is acquired. When voice is output from the voice analysis unit 114, a different call ID for each call, a group ID that is updated each time the speaker is switched, a sequential ID that is updated for each sentence of the speaker utterance, and the voice of the speaker Then, data is added to which is added a type ID that is a flag indicating whether it is text, translated speech, or text, and language ID information that represents the language of the speech, and is stored in the call data table 330 via the database access unit 160. Also, after the end of one sentence, the sequential ID is updated to accept the next sentence. The call ID, group ID, type, language ID, and sequential ID will be described later in the description of the call data table 330.

  As shown in FIG. 3, the transmission control unit 120 includes an encoder 121 and a call data input unit 122, and is a part that performs control for returning the translated voice to the telephone terminal 10 as a voice signal. The transmission control unit 120 performs control to extract voice data from the call data table 330 and send it back to the telephone terminal 10 in accordance with an instruction from the communication control unit 150. That is, at the timing instructed by the communication control unit 150, the call data input unit 122 of the transmission control unit 120 sets unsent voice data among the voice data in the specified language based on information such as a group ID and a sequential ID. The data is acquired from the call data table 330 via the database access unit 160, encoded by the encoder 121, and transmitted to the telephone terminal 10.

  The communication control unit 150 is a part that gives instructions to the language processing unit 200 and the transmission control unit 120 according to the analysis information of the DTMF signal sent from the reception control unit 110 and the state stored in the call data table 330.

  As shown in FIG. 3, the language processing unit 200 includes a speech recognition unit 210, a translation engine 220, and a speech synthesis unit 230. The language processing unit 200 inputs speech, translates it according to a designated language, and translates the speech of the translated language. As the output part. The voice recognition unit 210 recognizes the designated language type and converts it into text. The translation engine 220 translates a language (for example, Japanese) into another language (for example, English) based on the translation dictionary. The voice synthesizer 230 converts the text in the translation language into voice data and outputs it as one readable voice data.

  The database access unit 160 is a part that provides functions for reading and writing the database 300 from other components.

  The database 300 holds a button correspondence table 310, a call state table 320, and a call data table 330. Each table will be described in detail later.

  Each function of the communication server 100 may be implemented by hardware logic such as a field-programmable gate array (FPGA), or a general-purpose CPU (Central CPU) as a program loaded on a memory and operating on the OS. The function may be realized by the processing unit) executing the program.

Next, states and events handled on the communication server 100, and the relationship between them will be described with reference to FIGS. 4A to 6.
FIG. 4A is a diagram illustrating a state recognized by the communication control unit.
FIG. 4B is a diagram illustrating an event received by the communication control unit.
FIG. 5 is a state transition diagram showing state transition by event.
FIG. 6 is a diagram showing a processing matrix according to states and events that occur.

  The communication control unit 150 recognizes four states: “language not selected”, “acceptance”, “under translation”, and “translated speech transmission”. The meaning of each state is as shown in FIG. 4A. In the following figures, this number will be used to describe the state.

  The communication control unit 150 accepts five events of “DTMF signal (language selection)”, “voice”, “translation completion”, “translation voice transmission end”, and “DTMF signal (repeat playback)”. The meaning of each event is as shown in FIG. 4B.

  The state described above transitions when an event occurs. FIG. 5 shows the state and main things that occur at that time. For example, when a “DTMF signal (language selection)” event occurs in the “2: reception” state, the state transitions to “3: Under translation”, and when a “DTMF signal (repeat)” event occurs, When “4: Translated voice transmission is in progress” and a “voice” event occurs, it means that the state remains “2: reception”.

  6 is a matrix in which the state is represented by a column and the event is represented by a row. When an event represented by a row occurs in a state represented by a column, It shows that the portion corresponding to the intersection is applied. The element at the intersection is expressed as “state (accompanying process)”. When an event expressed in row occurs in the state expressed in the column, the state transitions to that state, and the accompanying process It shows that it starts or continues according to the transition.

  For example, when a “DTMF signal (language selection)” event occurs in the “2: reception” state, the process transits to “3: translation in progress”, a “translation start” process is performed, and a “DTMF signal (repeat)” When an “event” occurs, a transition is made to “4: Translation voice transmission in progress”, “Voice transmission” is started, and when a “Voice” event occurs, the state remains “2: Accept”, and “Voice data storage” ”Indicates that the process is continued.

  In the processing matrix of FIG. 6, the relationship between exceptional states and events not shown in the state transition diagram is also shown. However, in the description of the subsequent processing, it is mainly shown in the state transition diagram of FIG. Let's take a look at examples of events and events that occur.

Next, a data structure used in the communication system will be described with reference to FIGS.
FIG. 7 is a diagram for explaining the specifications of the button correspondence table 310.
FIG. 8 is a diagram for explaining the specifications of the call state table 320.
FIG. 9 is a diagram for explaining the specifications of the call data table 330.

  As shown in FIG. 7, the button correspondence table 310 has buttons # 1, language or function # 2, and language ID # 3. The button corresponding to the button pressed by the user determined from the DTMF signal and various information are displayed. It is a table to tie. The field of the button # 1 stores a button pressed by the user when the DTMF signal is generated by analyzing the DTMF signal. The language or function # 2 field stores the language or function corresponding to the value of the button # 1. For example, [#] [1] is Japanese, and [#] [*] is repeat. The language ID # 3 field stores a language ID corresponding to each language.

  As shown in FIG. 8, the call state table 320 has fields for call ID # 1, state # 2, and language ID # 3, and is a table that stores the state of processing referred to by the communication control unit 150. The call ID # 1 field stores a call identifier uniquely given to each call. The field of state # 2 stores an identifier representing the current state of the call identified by the call ID. The meaning of the state is as described in FIG. 4A. The language ID # 3 field stores the language ID of the currently selected language.

  As shown in FIG. 9, the call data table 330 has a field of call ID # 1, m (m is an integer of 0 or more) text # 10, and m speech # 20 structures. This is a table for storing information related to a call for each call ID. The call ID # 1 field stores a call identifier uniquely given to each call.

  The structure of text # 10 has members of type # 11, group ID # 12, language ID # 13, sequential ID # 14, and text data # 15, and information regarding the recognized text of the speaker and the translated text. Is stored. Type # 11 stores an identifier indicating whether the text is the speaker's recognized text or the translated text. The group ID # 12 stores a group ID that is uniquely assigned for each change of speaker. Language ID # 13 stores the language ID of the language of the text. Sequential ID # 14 stores a sequential ID assigned sequentially for each unit of audio data division. Text data # 15 stores text code data.

  The structure of the voice # 20 has members of a group ID # 21, a type # 22, a language ID # 23, a sequential ID # 24, and a voice data # 25. Stores information about data.

  The contents of group ID # 21, type # 22, language ID # 23, sequential ID # 24, and text data # 25 are the group ID # 11, type # 12, and language ID # 13 of the structure of text data # 10, respectively. This is the same as Sequential ID # 14. Audio data # 25 stores code data of audio data.

Next, an outline operation of the communication system will be described with reference to FIGS. 10A and 10B.
10A and 10B are schematic sequence diagrams showing delivery between components of the communication system and the system status.

  First, the speaker A (Japanese speaker) (SP1) presses the button ([#] [1]) on the telephone terminal 10, and transmits a DTMF signal to the communication server 100 side (A01). The state at this time is a language non-selected state (state = 1), and the values of the call state table 320 are call ID = 1, state = 1, language ID = 0 (T01) (S01, T01). Here, it is assumed that the call between the telephone terminal 10 and the communication server 100 is connected before A01, and the call ID has already been allocated. Further, the values of the call state table 320 and the call data table 330 are shown only by picking up values necessary for explanation.

  The communication control unit 150 of the communication server 100 receives the DTMF signal and updates the state and language (A20, (state = 2 (acceptance), language ID = 1 (Japanese): T02)).

In response to the DTMF signal (language selection) event, the system state transitions to reception (state = 2) (S02).
Next, it is assumed that voice data (Japanese) of the speaker A is transmitted from the telephone terminal 10 (A02). At this time, values are set in the structure of the voice # 20 in the call data table 330 (A21, group ID = 1, type = 0 (speaker), language ID = 1, sequential ID = 1, voice data: T03).

  Next, it is assumed that a call is handed over from Japanese speaker A to English speaker B (SP1 → SP2).

  Then, the speaker B (English speaker) (SP2) presses the button ([#] [2]) on the telephone terminal 10, and transmits a DTMF signal to the communication server 100 side (A03).

  Upon receiving the DTMF signal, the communication control unit 150 of the communication server 100 updates the state and language (A22, (state = 3 (under translation), language ID = 2 (English): T04)).

  Then, translation from Japanese into English is started, and a translation state (state = 3) is entered (S03).

Upon receiving an instruction from the communication control unit 150, the language processing unit 200 reads and translates the speech data in the call data table 330, writes the translated speech as new structure data, and when the translation is completed, the communication control unit 150 Then, the state of the call state table 320 is rewritten to “translated voice transmission (state = 4)” (A23, A24, T05, S04).
Next, the transmission control unit 120 extracts the translated speech (A25), and transmits it to the telephone terminal 10 as a translation result (Japanese → English) of the speech of the speaker A (A04).

  When the transmission of the translated voice is completed, the communication control unit 150 sets the state of the call state table 320 to the acceptance state (state = 2) (A26, S05).

  Here, it is assumed that the speaker B hears the voice transmitted through the telephone or feels that it is difficult to understand and has an intention to listen again. At this time, the speaker B operates the key device 13 of the telephone terminal 10 and presses a button ([#] [*]) instructing repeat. Thereby, the DTMF signal meaning repeat reproduction is transmitted from the telephone terminal 10 to the communication server 100 (A05).

  Then, the communication control unit 150 changes the state of the call state table 320 to a state in which translated speech is being transmitted (state = 4) (A27, S06).

  Next, according to the instruction of the communication control unit 150, the transmission control unit 120 extracts the translated voice data (A28), and again transmits the telephone terminal as a translation result (Japanese → English) of the speaker A. 10 (A06).

  When the second transmission of the translated voice is completed, the communication control unit 150 sets the state of the call state table 320 to the reception state (state = 2) (A29, S07).

  Next, it is assumed that the speaker B speaks and voice (English) is transmitted from the telephone terminal 10 (A07, A08).

  Thereby, the voice data is sequentially written in the call data table 330 (A30, T09, A31, T10).

  Next, it is assumed that the telephone is delivered from the English speaker B to the Japanese speaker A (SP2 → SP3).

  Then, the speaker A (Japanese speaker) (SP3) presses the button ([#] [1]) on the telephone terminal 10, and transmits a DTMF signal to the communication server 100 side (A09 in FIG. 10B).

  The communication control unit 150 of the communication server 100 receives the DTMF signal and updates the state and language (A32, (state = 3 (under translation), language ID = 1 (Japanese): T11)).

  Then, translation from English into Japanese is started, and a translation state (state = 3) is entered (S08).

Upon receiving an instruction from the communication control unit 150, the language processing unit 200 reads and translates the speech data in the call data table 330, writes the translated speech as new structure data, and when the translation is completed, the communication control unit 150 Then, the state of the call state table 320 is rewritten to “translated voice transmission” (state = 4) (A33, A34, T12, S09).
Next, the transmission control unit 120 sequentially extracts the translated voice data (A35) and transmits it to the telephone terminal 10 as a translation result (English → Japanese) of the voice of the speaker B (A10, A11). .

  When transmission of the translated voice is completed, the communication control unit 150 sets the state of the call state table 320 to the reception state (state = 2) (A36, S10).

  Next, it is assumed that the speaker A speaks and voice (Japanese) is transmitted from the telephone terminal 10 (A12).

  Thereby, voice data is written in the call data table 330 (A37, T14).

  Thereafter, the same sequence is repeated until the end of the call.

Next, processing of the communication system will be described with reference to FIGS. 11 to 13.
FIG. 11 is a schematic flowchart showing processing of the communication server.
FIG. 12 is a flowchart showing the voice analysis process.
FIG. 13 is a flowchart showing the repeat process.

  First, the communication server 100 receives a signal from the telephone terminal 10 via the telephone line 5, and the decoder 111 of the reception control unit 110 analyzes the transmitted signal (S100), and determines whether the signal is a control signal, a DTMF signal, or a voice. It is determined whether the signal is a signal, and is distributed to the control signal analysis unit 112, the DTMF signal analysis unit 113, and the voice analysis unit 114 according to the result (S101).

  When the control signal is a call signal, the control signal analysis unit 112 of the reception control unit 110 contacts the call data output unit 115 and sets a call ID (S130).

  When the control signal is a DTMF signal, the DTMF signal analysis unit 113 of the reception control unit 110 communicates the analysis information to the communication data output unit 115, and the communication data output unit 115 stores the DTMF signal according to the button correspondence table 310. Necessary language and function information is extracted (S102).

  When the control signal is an audio signal, an audio analysis process is performed (S150). The voice analysis process will be described later in detail with reference to the flowchart of FIG.

  When the control signal is the DTMF signal and the transmission of the DTMF signal is the first process (S103: Yes), the call data output unit 115 initializes the value of the group ID (S140) and sets the language ID to the call data table 330. (S141).

  When the transmission of the DTMF signal is not the first time (S103: No) and the DTMF signal is a signal indicating repeat (S104: Yes), the DTMF signal analysis unit 113 contacts the communication control unit 150. The transmission control unit 120 performs a repeat process in accordance with an instruction from the communication control unit 150 (S160), and the audio data extracted by the repeat process is encoded by the encoder 121 of the transmission control unit 120 into a transmission signal, so that the telephone terminal 10 Send to. The repeat process will be described later in detail with reference to the flowchart of FIG.

  If the DTMF signal is not a signal representing repeat (S104: No), the setting is switched to the language ID of the designated language (S105).

  Parameters such as a call ID, a language ID, a group ID, a type, and a sequential ID are passed from the call data output unit 115 to the language processing unit 200 via the communication control unit 150. The language processing unit 200 According to the instruction, the corresponding voice data is acquired from the call data table 330 (S106), and translation processing is performed (S107). Further, the language ID is switched, the type is the translated voice, and the voice data of the translated voice is stored in the call data table 330 (S108). Note that the text in which the voice data is recognized and the translated text are also written in the call data table 330.

  If all the sequential ID voice data (one sentence voice data) has been translated (S109: Yes), the process proceeds to the next step S110. If not translated (S109: No), the sequential ID is updated (S109: No). S113), returning to S106, the process is repeated.

  When all voice data belonging to the group ID is translated, the communication control unit 150 receives an instruction, and the call data input unit 122 of the transmission control unit 120 translates the voice data of the translated voice that has been translated from the call data table 330. (S110), and the call data output unit 115 of the reception control unit 110 updates the group ID (S111).

  Then, the encoder of the transmission control unit 120 encodes the audio data extracted in S110 into a transmission signal (S112) and transmits it to the telephone terminal 10.

  Next, the voice analysis process of S150 will be described with reference to FIG.

  First, the call data output unit 115 of the reception control unit 110 initializes a sequential ID (S200).

  Next, the presence / absence of voice data is determined (S201). When there is no voice data (S201: No), the process is terminated. When voice data is present (S201: Yes), the process goes to S202 (S202). ).

  When there is a break (silent part) in the voice data (S202: Yes), the voice data up to the break part is stored in the call data table 330 based on the call ID, language ID, group ID, and sequential ID ( S203), the sequential ID is updated (S204), and the process returns to the determination of S201.

  When there is no break in the audio data (S202: No), the break determination pointer is incremented (S205), and the process returns to the determination of S202.

  Next, the repeat process of S160 will be described with reference to FIG.

  First, the call data input unit of the transmission control unit 120 receives an instruction from the communication control unit 150 and transmits it from the call data table 330 immediately before based on the call ID, language ID, group ID, type, and sequential ID. Audio data is acquired (S300).

  When the audio data of all the sequential IDs in the group ID are acquired (S301: Yes), the process is ended. When there is audio data of the sequential ID that has not been acquired (S301: No), the sequential ID is changed. Update (S302) and return to S300.

The communication system according to the present embodiment is not an apparatus that generates a special signal, but is characterized in that it can be used by all telephone terminals that support the DTMF signal. Pushing a button on a phone is a widely used method, so even those who have never used automatic translation can easily use it without difficulty. In addition, the speaker clearly specifies the language, so There is a feature that the load on the side device is small and the determination can be performed in a short time.

  Also, in the face-to-face speech interpretation, when listening to the voice of automatic translation while alternately passing one phone, the translated voice of the other party is played after the user specifies his / her language type with the key. There is a feature that it is possible to prevent missed listening.

DESCRIPTION OF SYMBOLS 5 ... Telephone line 10 ... Telephone terminal 11 ... Handset 12 ... Transmitter 13 ... Key device 14 ... Voice signal conversion part 15 ... Connection control part 16 ... DTM signal generation part 100 ... Communication server 110 ... Reception control part 120 ... Transmission control Unit 150 ... Communication control unit 160 ... Database access unit 200 ... Language processing unit 300 ... Database 310 ... Button correspondence table 320 ... Call state table 330 ... Call data table

Claims (4)

A communication system in which a telephone terminal and a communication server are connected by a telephone line,
Means for transmitting and receiving a call to the communication server; means for generating a DTMF (Dual-Tone Multi-Frequency) signal by input from a key device; and transmitting the signal to the communication server via the telephone line; Have
The communication server includes a language processing unit that performs speech translation from a first language to a second language, and voice in a first language related to a voice signal transmitted through the telephone line. Means for translating and transmitting to the telephone terminal; voice data in the language of the speaker; and a call data table for storing voice data of translated voice obtained by translating the voice data;
When the communication server receives the first DTMF signal from the telephone terminal, it receives the first DTMF signal until it receives the second DTMF signal operated and transmitted by another speaker. Storing the voice related to the voice signal received in the call data table as voice data in the language represented by the first DTMF signal,
After the second DTMF signal is transmitted, the speech signal related to the speech data of the translated speech obtained by translating the speech data of the language represented by the first DTMF signal into the language represented by the second DTMF signal, A communication system characterized by transmitting to the telephone terminal.   The said communication server transmits again the audio | voice signal which concerns on the audio | voice data transmitted to the said telephone terminal after reception of the last DTMF signal, when the DTMF signal showing a repeat function is received. Communication system. A communication method in a communication system in which a telephone terminal and a communication server are connected by a telephone line,
The telephone terminal transmitting and receiving a call to the communication server;
The telephone terminal generates a DTMF (Dual-Tone Multi-Frequency) signal by input from a key device and transmits the signal to the communication server via the telephone line;
When the communication server receives the first DTMF signal from the telephone terminal, it receives the first DTMF signal until receiving the second DTMF signal operated and transmitted by another speaker. Storing the voice related to the voice signal received in the speech data table as voice data in the language represented by the first DTMF signal;
After the communication server transmits the second DTMF signal, the speech data of the language represented by the first DTMF signal is converted into the speech data of the translated speech that is translated into the language represented by the second DTMF signal. And transmitting the voice signal to the telephone terminal.   The communication server includes a step of retransmitting an audio signal related to audio data transmitted to the telephone terminal after receiving the immediately preceding DTMF signal when receiving a DTMF signal representing a repeat function. 3. The communication method according to 3.

Images