JP2007243392A - Call terminal, multi-user call system, multi-user call method, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a call terminal which assures shifting of the right to speak while a group session is established for smooth communication. <P>SOLUTION: The call terminal allows multi-user call. It comprises a received voice recognizing means which receives voice data transmitted from a call terminal of a user having the right to speak to recognize voice, a speech request detecting means which detects that the received voice data indicate speech request based on the voice recognition result, and a reporting means which, when detecting a speech request, reports it to a user. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a call terminal, and more particularly to a call terminal capable of multi-party calls between a plurality of terminals.

  In recent years, with the improvement of communication technology in mobile phones and the like, a group session is established between a plurality of terminals, and one-to-N (many) calls are possible. Such a technique is called by the name of Push-to-Talk over Cellular (PoC) in OMA (Open mobile Alliance) which is a mobile service standardization organization.

  The above-described PoC communication with a plurality of people is performed by a half-duplex communication method close to a transceiver. That is, while a person who has acquired the right to speak speaks, other persons become listeners. The right to speak is exchanged between the members of the PoC communication group, and it is common practice to encourage other participants to obtain the right to speak by verbal cues (request to speak) such as is there.

JP 2001-188740 A

  However, in the above-described conventional verbal request acquisition request only by oral, if the request for speech cannot be heard due to ambient noise or if it is not heard because of distraction, the signal (speak request) is not transmitted smoothly. May hinder proper communication. In addition, when urging people to speak in PoC communication group members, it is not only a verbal request for speech, but at the same time, it can be discussed if notification can be encouraged by using more effective notification means. It is thought that it will become more active.

  On the other hand, in Patent Document 1, in a teleconference system, by pressing a button for designating a speech target person, the right to speak is given to the target person, and the target person is notified accordingly. Is disclosed. However, in order to facilitate smooth communication in the group session, it is desirable to encourage other participants to obtain the right to speak by verbal cues in the conversation, as in the conventional example.

  For this reason, the present invention improves the inconvenience of the above-described conventional example, and in particular, provides a call terminal capable of improving the certainty of floor transfer during establishment of a group session and achieving smooth communication. Is the purpose.

Therefore, a call terminal according to an aspect of the present invention is
A call terminal capable of multi-party calls,
A received voice recognition means for receiving and recognizing voice data transmitted from a call terminal of a person who has the right to speak;
A speech request detecting means for detecting that the speech data received based on the speech recognition result is a speech request;
An informing means for informing the user when a request to speak is detected;
It is characterized by having.

  According to the above invention, first, when a voice requesting speech is uttered from a person who has the right to speak, the speech requesting voice is automatically detected by voice recognition at the call terminal of the user who has requested speech. Then, the user is notified of the request for speech by a ringing tone or a vibrator at the call terminal. Therefore, the user can recognize that the request for speech has been made not only by the voice in the call but also by other methods, and smooth communication can be achieved.

  The speech request message storage unit stores a speech request message set in advance, and the speech request detection unit detects a speech request based on the stored speech request message. As a result, it is possible to detect the speech request wording with high accuracy and achieve smoother communication.

  Further, the notification means is characterized in that a plurality of types of notification are performed, and in particular, the notification means is characterized in that a plurality of types of notification are performed at different times. Furthermore, it is provided with a speech request detection frequency count means for counting the number of speech request detections within a certain time, and the notification means performs a plurality of types of notification according to the speech request detection frequency. Thereby, while being able to perform effective alerting according to the call status, it is possible to suppress the alerting from interfering with the user's phone call, and further smooth communication can be achieved.

Moreover, the other form of the call terminal in the present invention is:
A speech request speech recognition means for recognizing a speech by receiving an input of a speech request to another user;
A synthesized sound generating means for generating a synthesized sound corresponding to the speech-recognized voice data;
A speech request transmitting means for transmitting the voice data composed of the generated synthesized sound to a call terminal of another user,
It is characterized by having.

  According to the above invention, the voice of the speech request issued from the person who has the right to speak is recognized by the call terminal and transmitted to the other user's call terminal as a synthesized sound. Accordingly, since a speech request is made with a synthetic sound that is easy to hear, users of other call terminals can easily recognize the speech request, and smooth communication can be achieved. In particular, as described above, when a speech request is automatically detected by voice recognition at the call terminal of the user who has requested speech, the speech request can be detected with higher accuracy.

  And it is provided with the speech request message memory | storage means which memorize | stores the speech request message set beforehand, and a synthetic sound production | generation means produces | generates the synthetic | combination sound corresponding only to specific audio | voice data based on the stored speech request wording. It is characterized by that. As a result, a synthesized sound of only the speech request message is generated, so that the processing burden on the terminal can be reduced.

  In addition, voice feature data storage means for storing voice feature data representing the voice characteristics of the operator in advance is provided, and the speech request voice recognition means performs voice recognition based on the voice feature data. As a result, it is possible to detect the speech request wording with high accuracy according to the operator's habit and to achieve smoother communication.

  The present invention also provides a multi-party call system including a call terminal owned by a person who has the right to speak and makes a request for speech, and a call terminal of a user who receives the request for speech. Further, the above-described call terminal is configured by a single terminal.

Furthermore, another aspect of the present invention is:
A server computer connected via a network to a call terminal capable of multi-party calls,
A speech request speech recognition means for recognizing speech by receiving a speech request for speech sent to another user by a server computer from a call terminal of a person who has a speech right;
A speech request detecting means for detecting that the speech data received based on the speech recognition result is a speech request;
Notification control means for notifying other users' call terminals when a request to speak is detected;
It is characterized by having.

  Thereby, since the voice recognition processing can be executed by the server computer, the above-described effects can be obtained, and the processing load on the call terminal can be reduced.

Moreover, the program which is the other form of this invention is:
In the computing device equipped in the call terminal that can make multi-party calls,
A received voice recognition means for receiving and recognizing voice data transmitted from a call terminal of a person who has the right to speak;
A speech request detecting means for detecting that the speech data received based on the speech recognition result is a speech request;
Notification control means for notifying the user via a notification device equipped in the call terminal when a request to speak is detected;
It is a program for realizing.

Another form of the program according to the present invention is as follows:
In the computing device equipped in the call terminal that can make multi-party calls,
A speech request speech recognition means for recognizing a speech by receiving an input of a speech request to another user;
A synthesized sound generating means for generating a synthesized sound corresponding to the speech-recognized voice data;
A speech request transmitting means for transmitting the voice data composed of the generated synthesized sound to a call terminal of another user,
It is a program for realizing.

Furthermore, the present invention provides a multi-party call method using a call terminal,
The user's call terminal that receives the request for speech receives the voice data transmitted from the call terminal of the person who has the right to speak and recognizes the voice, and the voice data received based on the voice recognition result A speech request detection step for detecting that it is a request, and a notification step for notifying the user via a notification device equipped in the communication terminal when a speech request is detected,
It is characterized by having.

  The speech request detection step is characterized in that a speech request is detected based on a speech request message stored in advance in the speech request message storage means. Further, the notification step is characterized in that a plurality of types of notification are performed at different times. Further, the notifying step is characterized in that the number of times of utterance request detection within a predetermined time is counted and a plurality of types of notifications are performed according to the number of utterance request detections.

  Further, prior to the received voice recognition step, the speech terminal of the person who has the right to speak receives a speech request input by voice to another user and recognizes the voice, and this voice-recognized voice data A synthesized sound generating step for generating a synthesized sound corresponding to the voice, and a speech request transmitting step for transmitting voice data composed of the generated synthesized sound to a call terminal of another user.

  The synthesized sound generation step is characterized in that a synthesized sound corresponding to only specific voice data is generated based on the speech request message stored in advance in the speech request message storage means. The speech request speech recognition step is characterized in that speech recognition is performed based on speech feature data representing an operator's speech features stored in advance in speech feature data storage means.

  Even the multi-party call system, the program, and the multi-party call method configured as described above operate in the same manner as the above-described call terminal, so that the above-described object of the present invention can be achieved.

  Since the present invention is configured and functions as described above, according to this, since a speech request voice from another person is automatically detected by voice recognition, the user cannot hear a speech request in a call. Even in this case, it is possible to recognize that there is a request for speech by another notification method. Therefore, the present invention has an unprecedented excellent effect that smooth communication during multi-party calls can be achieved.

  The present invention is characterized in that the certainty of recognition of voice data, which is a speech request for encouraging the transfer of the right to speak, is improved in order to facilitate smooth communication during a multi-party call. Hereinafter, in the embodiment, a mobile phone is given as an example of a call terminal capable of multi-party calls, and Push-to-Talk over Cellular (PoC) is established by OMA (Open mobile Alliance) which is a mobile service standardization organization. A case where a multi-party call is performed using a technique called “name” will be described as an example. However, the call terminal is not limited to a mobile phone, and a system for performing multi-party calls is not limited to the PoC system.

  A first embodiment of the present invention will be described with reference to FIGS. FIG. 1 is a schematic diagram showing an overall configuration of a multi-party call system, and FIG. 2 is a block diagram showing a configuration of a mobile phone that realizes multi-party call. FIG. 3 is a diagram illustrating an example of information stored in the mobile phone. FIG. 4 is an explanatory diagram showing a state during a multi-party call, and FIGS. 5 to 6 are sequence diagrams showing an operation during a multi-party call.

[Constitution]
As shown in FIG. 1, the multi-party call system in the present invention is realized by mobile phones a to d operated by a plurality of users A to D. In particular, in the present embodiment, the respective mobile phones a to d are connected via the network N, and a multi-party call is performed by the PoC communication described above. In the multi-party call in this embodiment, the voice data from the mobile phone (for example, a) of the user (for example, A) who has the right to speak is transmitted from other users (listeners) (for example, BD). This is realized by being transmitted to a mobile phone (for example, b to d). And in order for another user to speak, it is necessary to obtain the right to speak, and a speech is made after receiving a speech request from a user having the right to speak.

  Hereinafter, a case where the user A (speaker) has the right to speak and the user B (listener) receives a speech request will be described as an example. That is, a case will be described in which a speech request is transmitted from the cellular phone with the symbol a to the cellular phone b with the symbol b. Note that any of the mobile phones a to d is configured in the same manner, and any user A to D can be a person who has the right to speak or a listener.

  FIG. 1 shows a configuration of a mobile phone. Here, a mobile phone with the symbol a will be described as an example, but the other mobile phones b to c have the same configuration. As shown in this figure, the mobile phone a is connected to each main part in the apparatus via an internal bus 13 with the apparatus control section 6 as the center. The device control unit 6 includes a CPU (Central Processing Unit), a cache memory, an interrupt controller, a DSP (Digital Signal Processor), and an OS for controlling the entire device, and has a role of controlling the entire portable terminal device.

  And as a device (notifier) for notifying a user (listener) of a speech request, a vibration drive device portion 1, an LED portion 2, a display portion 3, and a main portion in the device equipped in a general mobile phone, The speaker unit 4 is applicable. The vibration drive unit 1 is equipped with a mechanism that vibrates the entire apparatus. When a request for speech is received, the vibration drive unit operates to notify the user. The LED unit 2 is disposed on the front surface of the apparatus main body from which the user can confirm the light emission. When the LED request is received, the LED emits light to notify the user. The display unit 3 is a display device having a display control function. The display device is a display device such as a liquid crystal panel or organic EL (electrofluorescence). When a request for speech is received, the display on the entire screen or a part of the screen changes to notify the user. As the image data to be displayed at this time, user data stored in a nonvolatile area (ROM) of the memory 5 is used. The speaker unit 4 is a device that outputs a voice conversation during a PoC call, and at the same time, is used for a speech request notification unit. When there is a request for speech, a ringing tone designated in advance is sounded to notify the user. As a ringing tone, not only a mechanical tone but also music / voice data stored in a nonvolatile area (ROM) of the memory 5 may be used.

  In addition, operation | movement of the vibration drive apparatus part 1, LED part 2, the display part 3, and the speaker part 4 which are the alerting devices mentioned above is controlled by the apparatus control part 6 as an alerting | reporting control means mentioned above. In particular, as will be described later, when a request for speech is detected by the mobile phone b of the listener B, control is performed so that a notification operation is performed by a notification device. Therefore, the notification unit is configured by the notification device and the device control unit 6.

  The memory 5 is equipped with both ROM / RAM. The ROM is a non-volatile memory that stores user data such as a telephone directory and music / image data. The RAM stores data temporarily required for the CPU to execute the program. The memory 5 in the present embodiment stores a speech request keyword (speech request message) set in advance, and the memory 5 functions as a speech request message storage unit. An example of the speech request keyword is shown in FIG. As shown in this figure, the speech request keyword is stored as words of a certain delimiter, and is used by combining each word by a logical operation (AND, OR, NOT, etc.). For example, in the case of an utterance request such as “Mr. A, please speak,” voice recognition processing or utterance request detection described later using the result of the AND operation of tables No. 1 and No. 4 in FIG. Processing is performed. If the request is “a person with few utterances, please speak”, the time measured by the timer unit 12 is used to preliminarily measure the utterance time for each speaker, and tables No. 4 and No. 4 in FIG. The result of the 6 AND operation is used. These speech request keywords are used at the time of speech request in the mobile phone a of the speaker A, and are used when detecting the speech request in the mobile phone b of the listener.

  The transmission control unit 7 is a circuit that performs transmission control of voice data and the like wirelessly via the communication antenna 9. The reception control unit 8 is a circuit that performs reception control of voice data and the like wirelessly via the communication antenna 9.

  The voice recognition unit 10 (speech request voice recognition unit, received voice recognition unit) is a voice at the time of a call such as a call data input from the reception part or a call data received from another mobile phone via the reception control unit 8. It is realized by a device or a program that executes recognition / detection of specific keywords or specific control sounds with respect to data by a predetermined detection algorithm. Details of the speech recognition algorithm are not mentioned in the present invention as known ones, but when used by the user A who has the right to speak, learning the individuality of the owner to improve the speech recognition accuracy Is possible. That is, voice feature data representing the voice feature of the user is stored in advance in a ROM or the like (voice feature data storage means), and voice recognition is performed based on this.

  The voice recognition unit 10 (speech request detecting means) constantly operates during his / her speech and when other participants speak, and has a role of detecting a keyword that becomes a speech request included in the conversation. At this time, a speech request keyword as shown in FIG. 3 stored in the ROM described above is detected. Then, when used on the side of the listener B, when the speech recognition device unit 10 detects a speech request, as described above, the vibration driving device unit 1 or the LED unit 2 via the device control unit 6, or The user is notified from the display unit 3 or the speaker unit 4.

  The synthesized sound generating unit 11 (synthesized sound generating means) is a device that can process and output input voice data. In the present invention, in particular, it operates on the mobile phone a on the transmitting side, that is, on the mobile phone a of the user A who is the PoC speaker side, and the voice recognition unit 10 of the telephone terminal b on the receiving side (PoC listener side). This is to improve the voice recognition accuracy. Specifically, as described above, when a speech request keyword is detected from the voice data input from the reception part, a synthesized sound of the keyword itself is generated, and another call terminal via the transmission control unit 7 Send to. That is, in cooperation with the transmission control device 7, it also functions as a speech request transmission unit. Further, the synthesized sound generating unit 11 adds a specific identification sound assigned to each of the speech requesting parties, changes an intonation (for example, changes a local round intonation into a standard Japanese intonation), and the like. Processing.

  The timer unit 12 is a time measuring device for measuring time. As will be described later, it is used in a time measurement application within a detection request detection algorithm. The microphone unit 14 is a device that collects the content of a speaker's speech and is a receiving part.

[Operation]
Next, the operation of the mobile phone configured as described above will be described with reference to FIGS. In the following, first, an overview of the overall operation will be described with reference to FIG. 4, followed by description of the operation of the mobile phone a of the user A on the transmission side with reference to FIG. 5, and further with reference to FIG. The operation of the mobile phone b of the user B who is the receiving side will be described.

<Overview of overall operation>
FIG. 4 is an image diagram showing a state in which a PoC conference is performed using the mobile phones A to D described above. Here, a state where the user A is a speaker, the user B is a listener, and the user A requests the user B to speak is described.

  The PoC communication is established between the cellular phone a of the speaker (user A) and the cellular phone b of the listener (user B) in the figure, and the speaker A is now speaking. It is assumed that the content of the utterance includes a voice “Please speak, Mr. XX”, which is a utterance request to the listener B.

  Then, in the mobile phone a on the transmission side, the real voice of the speaker A is input as voice data (Y1). In the real voice data at this stage, there are many differences in the data even if the content is the same, for example, due to the manner in which the speaker speaks, the tone of the voice, and the roundness. This individual difference is a factor that causes a decrease in the voice recognition accuracy of the mobile phone b on the receiving side, and there is a problem that the notification operation malfunctions in an unexpected situation when the recognition accuracy is low. Therefore, in the present embodiment, as described above, the speech request part of the real voice data is detected using the speech recognition unit of the transmitting mobile phone a having learned the voice characteristics of the speaker A, and the speech request The voice data of the part is converted into synthesized sound (Y2). Then, the individual difference intervening in the real voice data is removed by the conversion to the synthesized sound, and an effect of improving the recognition accuracy in the mobile phone b on the receiving side can be expected. As a synthetic sound generation method, a specific identification sound may be added to the voice data, or the intonation change of the speech (removal of speech such as a round) may be performed.

  As described above, the voice data is synthesized in the speech request portion and transmitted to the receiving terminal device through the PoC communication network (Y3). Then, in the mobile phone b on the receiving side, the voice data (“Mr. XX, please say”) of the received synthesized sound is output to the outside by a ringing device such as a speaker (Y4), and the voice recognition unit also has a conversation. Continuously detect requests to speak. When the speech recognition unit detects a speech request that has been synthesized (Y5), a speech request notification operation is performed (Y6). In addition, the conversation output operation | movement of the code | symbol Y4 may be abbreviate | omitted and only the alerting | reporting operation | movement of the code | symbol Y6 may be performed.

<Operation of sending mobile phone>
Next, the operation in each of the mobile phones a and b will be described in detail. First, the operation of the cellular phone a of the speaker A will be described in detail with reference to the sequence diagram of FIG.

  The utterance by the speaker A is input from the microphone unit, converted into voice data (step S1), and sequentially transferred to the memory unit (step S2). The voice data is stored in a voice recognition buffer of the memory unit 5 (step S3), and the voice recognition unit 10 performs voice recognition processing from the voice data stored in the buffer (step S4). The voice recognition process always operates during a conversation and detects a specific keyword that is a request for speech (step S5).

  When a specific keyword is detected, the data of the voice part is transferred to the synthesized sound generating unit 11 (step S7), and converted into synthesized sound by the synthesized sound generating unit 11 (step S8). The data is transferred to the transmission control unit 7 (step S9). On the other hand, the voice data in which the specific keyword is not detected is transferred to the transmission control unit 9 as the real voice data as it is (step S6). The transmission control unit 7 encodes the transferred data and transmits the data from the communication antenna 9 to the mobile phone b of the listener B (step S10).

<Operation of receiving mobile phone>
Next, the operation of the mobile phone b of the listener B will be described in detail with reference to the sequence diagram of FIG. First, the device control unit 6 is always in a speech request detection waiting state (step S20). When the communication antenna 9 receives data from the above-described speaker A's mobile phone a, the reception control unit 8 decodes the voice data (step S21). Thereafter, the received voice data is sequentially transferred (step S22), and the conversation is output from the speaker unit 4 (steps S29 and S30), and stored in the voice recognition buffer of the memory unit 5 (step S23). The voice recognition unit 10 performs voice recognition processing from the voice data stored in the buffer (step S24). The voice recognition process always operates during a conversation, and a specific keyword that is a request for speech is detected (step S25).

  When the request to speak is detected, the specific keyword hit is immediately notified to the device control unit 6 (step S26), and the device control unit 6 receives the notification, and the speaker unit 4, the display unit 3, A notification operation request is made to the LED unit 2 and the vibration control unit 1 (step S27). Then, in response to the notification operation request, each unit 1 to 4 performs a notification operation (step S28). Note that the speech request notification operation may be performed simultaneously by each unit, or may be performed independently at different times.

  As described above, in the present embodiment, a speech request such as “Please speak ○○” during a PoC meeting is automatically detected by voice recognition on the receiving mobile phone b, and the vibration of the device, LCD display It is possible to operate notification means such as LED lighting and control sound response. That is, the user can obtain an opportunity to know the request for speech not only by voice but also by other operations of the terminal device. Therefore, by making sure that the request for speech is notified to the other party more reliably than before, the situation where the request for speech was not heard due to ambient noise or the request for speech was not heard due to distraction is eliminated, and smooth communication is achieved. be able to.

  Furthermore, in the present embodiment, voice recognition is also performed at the telephone terminal a on the transmission side, and the voice data of the recognized speech request part is converted into a synthesized sound, so that individual differences such as the manner in which the speaker speaks and the roundness are reduced. It can be generalized. Thereby, the voice recognition accuracy by the mobile phone b on the receiving side is improved, and further smoothing of communication can be achieved.

  Note that, unlike the above, the speech request voice from the mobile phone a on the transmission side may be transmitted to the mobile phone b on the reception side as a real voice without being synthesized. In this case, the mobile phone b on the receiving side performs voice recognition on the real voice and detects a speech request. Even in this case, since the speech request voice is automatically detected by voice recognition in the mobile phone b on the receiving side, smooth communication can be achieved.

  Also, unlike the above, voice recognition is not performed on the mobile phone b on the receiving side, and the speech request speech from the mobile phone a on the sending side is synthesized and transmitted only to the mobile phone b on the receiving side. May be. Even in this case, since the request for speech is made to the reception-side mobile phone b with a synthesized sound that is easy for a listener to hear (easy to recognize), smooth communication can be achieved.

  Note that the above-described configuration can be realized only by changing the configuration of the mobile phone, and does not require improvement of the current PoC communication system that enables multi-party calls, thus reducing the cost burden. .

  Next, a second embodiment of the present invention will be described with reference to FIG. FIG. 7 is a sequence diagram showing a notification operation of the mobile phone b on the receiving side.

  If a notification operation is performed due to misrecognition of voice recognition, communication may be hindered. In particular, user dissatisfaction is great if the vibration drive device operates. For this reason, as a control method for reducing the influence on the user due to misreporting, in this embodiment, the operation is performed step by step in order of increasing strength of the user notification. For example, as described above, if there are four notification means of vibration, LCD display, LED lighting, and sounding, control is performed in the order of LED lighting → LCD display → sounding → vibration.

  A specific operation will be described with reference to FIG. Here, the mobile phone b on the receiving side is in a waiting state for detecting the speech request from the mobile phone a on the transmitting side (step S40), the device control unit 6 receives the speech request, and the notification means operation unit 1; The informing means operating unit 2 is in a situation to be controlled. When the device control unit 6 receives the speech request (step S41), it makes a timer start request to the timer unit 12 (step S42), and the timer unit 12 performs timer measurement for a certain time (step S43). When the timer measurement is finished, the timer unit 12 notifies the device control unit 6 of the count-up completion (step S44). Therefore, for the first time, the device control unit 6 notifies the notification unit operation unit 1 of a notification operation request (step S45), and the notification operation is performed (step S46). At this time, the 1st time is performed using the alerting | reporting device with weak alerting | reporting strength with respect to the observer B, such as light emission by the LED part 2. FIG.

  Similarly, when the time is further measured and a certain time has elapsed, a second notification is performed as shown in steps S47 to S51. At this time, the notification is performed using a notification device having a higher notification degree such as notification by sound from the speaker unit 4.

  As described above, a plurality of types of notifications are performed at different times, and further, notification operations are performed due to misrecognition of voice recognition by performing notification in stages while changing the notification strength for the user. It operates in a fail-safe manner in some cases, and has the effect of reducing user dissatisfaction due to malfunction.

  Next, a third embodiment of the present invention will be described with reference to FIG. FIG. 8 is a sequence diagram showing the notification operation of the mobile phone b on the receiving side.

  The present embodiment is characterized in that notification is made by changing the notification method according to the number of detections of a speech request within a certain time. In other words, the device control unit 6 has a function (speech request detection frequency counting means) that counts the number of times that a speech request (a part of the keywords) is detected within a certain time, and various notifications according to the number of times. It has a function of performing. This is in consideration of the tendency that the request for speech is repeated in a short time. For example, if the listener B is “Mr. XX”, the first request to speak is “I want to receive a comment from Mr. XX”. "The third request to speak is assumed to be a conversation such as" Do you hear Mr. XX? ". In this case, the specific keyword for counting the number of detections is" Mr. XXX "

  As shown in FIG. 8, first, the device control unit 6 is in a first detection waiting (hit waiting) state (step S60). When notification of detection of the specific keyword is made (step S61), a timer start request is made to the timer unit 12 (step S62), timer measurement is started in the timer unit 12 (step S63), and the count of the specific keyword is counted. Be started. At the same time, the apparatus control unit 6 enters a second wait state (step S64). After that, when the device control unit is notified of the specific keyword hit (second time) while waiting for the second hit (step S65), first, a notification operation request is issued to the notification unit operation unit 1 which performs a relatively weak notification unit. (Step S66) is notified, and the notification means is operated (Step S67). Furthermore, when a specific keyword hit notification (third time) is made again (step S68), a notification operation request (step S69) is notified to the notification means operation unit 2 that performs strong notification means, and the notification means is activated (step S69). Step S70).

  As shown in the dotted line in FIG. 8, when the timer count-up is completed before the second or subsequent keyword hit is notified, the timer unit 12 notifies the device control unit 6 of the count-up completion notification. If this is done (step S71), the device control unit 6 transitions to the first hit waiting state (step S72).

  As a result, it is possible to notify the request for speech more reliably and to achieve further smooth communication.

  Next, a third embodiment of the present invention will be described with reference to FIG. In the above description, the speech recognition processing for requesting speech is performed by each of the mobile phones a and b. However, such processing may be executed by the PoC server system 20. In other words, the server system 20 includes the speech recognition unit 10 (speech request speech recognition means, speech request detection means) included in the mobile phone b described above, and further indicates that the speech request has been detected. The function (notification control means) of notifying is provided.

  The operation will be described with reference to FIG. First, when an utterance request “Mr. XX, please speak” is input from the speaker (A) to the listener (B), the mobile phone a on the transmitting side directly inputs the input message (Y11) to the PoC server. The data is transmitted to the system 20 (Y12, Y13). The PoC server system 20 includes a voice recognition unit, detects a speech request in the same manner as the telephone terminal b described in the above embodiment (Y14), and requests the speech request to the mobile terminal device of the person who is the request target. Is transmitted (Y15). At this time, the audio data is also transmitted as it is. Then, in the portable terminal b that has received this control signal, the speech request notification (Y17, Y18) operates during the conversation output (Y16).

  Thereby, since the voice recognition processing can be executed by the server system 20, the above-described effects can be obtained, and the processing burden on the mobile phones a and b can be reduced.

  Next, a fifth embodiment of the present invention will be described with reference to FIGS. 10 to 11 are diagrams illustrating a configuration of a mobile phone.

  The mobile phone in this embodiment is a mobile phone a on the transmission side that has a right to speak and makes a request to speak to another user. Basically, the configuration is the same as that of the mobile phones a and b in the above-described embodiment as shown in FIG. 2, but a keyboard portion 15 is further provided as shown in FIGS. Is different. As shown in FIG. 11, each key of the keyboard unit 15 is associated with a PoC conference participant and a control voice capable of discriminating the participant, and the association information is stored in advance in the control sound conversion table. 51 is stored in the memory unit 5. Such association is performed by the user A.

  Then, when there is a partner who wants to make a speech request, the speaker A presses a key corresponding to the speaker (Y21), and the corresponding control sound is generated by the synthesized sound generation unit 11. And since this control sound is individual for each participant, the telephone terminal b on the receiving side can detect that it is a speech request to another user B by voice recognition processing. That is, the control sound functions as a synthetic sound for the above-described speech request.

  In this way, the mobile phone a on the transmitting side can obtain the same effect as when a request for speech is made only by pressing a key, without requesting a statement of “Please say Mr. XX”. This function may be used in an auxiliary manner when issuing the above-described speech request by voice. That is, the control sound may be transmitted together with a request for speech by a synthesized sound generated based on the voice.

  The present invention has a communication function of a mobile terminal device such as a mobile phone, PHS (Personal Handyphone System), PDA (Personal Data Assistance, Personal Digital Assistants), and a PC (Personal Computer). Therefore, it can be used for a terminal device capable of multi-party calls and has industrial applicability.

It is the schematic which shows the whole structure of a multi-party call system. 1 is a block diagram illustrating a configuration of a mobile phone in Example 1. FIG. It is a figure which shows an example of the speech request keyword memorize | stored in the mobile telephone. It is explanatory drawing which shows the mode at the time of the multi-party call in Example 1. FIG. It is a sequence diagram which shows the operation | movement at the time of the multi-party call in Example 1. It is a sequence diagram which shows the operation | movement at the time of the multi-party call in Example 1. It is a sequence diagram which shows the operation | movement at the time of the multi-party call in Example 2. It is a sequence diagram which shows the operation | movement at the time of the multi-party call in Example 3. It is explanatory drawing which shows the mode at the time of the multi-party call in Example 4. FIG. 10 is a block diagram illustrating a configuration of a mobile phone according to a fifth embodiment. It is explanatory drawing which shows the mode at the time of the multi-party call in Example 5. FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Vibration drive device part 2 LED part 3 Display part 4 Speaker part 5 Memory 6 Device control part 7 Transmission control part 8 Reception control part 9 Communication antenna 10 Speech recognition part 11 Synthetic sound production | generation part 12 Timer part 13 Internal bus 14 Microphone Part 15 Keyboard part A User (speaker)
B user (listener)
a, b, c, d mobile phone

Claims (19)

A call terminal capable of multi-party calls,
A received voice recognition means for receiving and recognizing voice data transmitted from a call terminal of a person who has the right to speak;
A speech request detection means for detecting that the received speech data is a speech request based on the speech recognition result;
An informing means for informing the user when a request to speak is detected;
A call terminal characterized by comprising: A speech request message storage means for storing a speech request message set in advance;
The speech request detection means detects a speech request based on the stored speech request wording.
The call terminal according to claim 1.   The call terminal according to claim 1 or 2, wherein the notification means performs a plurality of types of notification.   The call terminal according to claim 3, wherein the notification means performs the plurality of types of notifications at different times. A speech request detection frequency counting means for counting the number of speech request detections within a predetermined time is provided,
5. The call terminal according to claim 3, wherein the notification unit performs the plurality of types of notifications according to the number of detected speech requests. A call terminal capable of multi-party calls,
A speech request speech recognition means for recognizing a speech by receiving an input of a speech request to another user;
A synthesized sound generating means for generating a synthesized sound corresponding to the speech-recognized voice data;
A speech request transmitting means for transmitting the voice data composed of the generated synthesized sound to the other user's call terminal;
A call terminal characterized by comprising A speech request message storage means for storing a speech request message set in advance;
The synthesized sound generating means generates a synthesized sound corresponding to only specific voice data based on the stored message request wording;
The call terminal according to claim 6. Voice feature data storage means for storing voice feature data representing the voice feature of the operator in advance,
The speech request speech recognition means performs speech recognition based on the speech feature data.
The call terminal according to claim 6 or 7, characterized by the above.   A multi-party call system comprising the call terminal according to claim 1 and the call terminal according to claim 6. A server computer connected via a network to a call terminal capable of multi-party calls,
A speech request speech recognition means for recognizing the server computer by receiving a speech request for speech transmitted to the other user from the call terminal of the person having the speech right;
A speech request detection means for detecting that the received speech data is a speech request based on the speech recognition result;
Notification control means for notifying the other user's call terminal of that when a request to speak is detected;
A server computer comprising: In the computing device equipped in the call terminal that can make multi-party calls,
A received voice recognition means for receiving and recognizing voice data transmitted from a call terminal of a person who has the right to speak;
A speech request detection means for detecting that the received speech data is a speech request based on the speech recognition result;
Notification control means for notifying the user via a notification device equipped in the call terminal when a request to speak is detected;
A program to realize In the computing device equipped in the call terminal that can make multi-party calls,
A speech request speech recognition means for recognizing a speech by receiving an input of a speech request to another user;
A synthesized sound generating means for generating a synthesized sound corresponding to the speech-recognized voice data;
A speech request transmitting means for transmitting the voice data composed of the generated synthesized sound to the other user's call terminal;
A program to realize A multi-party call method using a call terminal,
A speech terminal of a user who receives a speech request receives a speech data transmitted from a speech terminal of a person who has a speech right and recognizes the speech, and the received speech data is based on the speech recognition result. A speech request detection step for detecting that it is a speech request; a notification step for notifying the user via a notification device equipped in the communication terminal when a speech request is detected;
A multi-party call method comprising: The speech request detection step detects a speech request based on a speech request word stored in advance in the speech request word storage unit.
The multi-party call method according to claim 13.   The multi-party call method according to claim 13 or 14, wherein the notification step performs a plurality of types of notifications at different times. The notifying step counts the number of times a request for speech is detected within a certain period of time, and performs a plurality of types of notification according to the number of times a request for speech is detected.
16. The multi-party call method according to claim 13, 14, or 15. Prior to the received speech recognition step, a speech terminal of a person who has the right to speak receives a speech request input by speech to another user and recognizes speech, and the speech-recognized speech data A synthesized sound generating step for generating a corresponding synthesized sound, and a speech request transmitting step for transmitting voice data composed of the generated synthesized sound to the call terminal of the other user;
17. The multi-party call method according to claim 13, 14, 15 or 16, characterized by comprising: The synthetic sound generation step generates a synthetic sound corresponding to only specific voice data based on a speech request word stored in advance in the speech request word storage unit.
The multi-party call method according to claim 17. The speech request voice recognition step performs voice recognition based on voice feature data representing an operator's voice feature stored in advance in the voice feature data storage means.
The multi-party call method according to claim 17 or 18, characterized in that:

Images