KR101074069B1 - System of Private Branch Exchange and Method of Speech Synthesis - Google Patents

Abstract

According to an embodiment of the present invention, a method of synthesizing a voice for a announcement by a speech synthesis resource server in association with a resource management server includes: exchanging state information between a resource management server and a speech synthesis resource server; When the available channel information is requested from the call box, the resource management server identifying an available channel of the voice synthesis resource server and transmitting server IP information of the available available channel to the call box; And receiving a speech synthesis request including sentences corresponding to the server IP information and announcement from the call box, the speech synthesis resource server generates a speech synthesis file for the sentence, and a result of the speech synthesis request. Transmitting to the call box.

On-premises exchange, speech synthesis, resources

Description

System of Private Branch Exchange and Method of Speech Synthesis

The present invention relates to an on-premises exchange system and speech synthesis method.

A private branch exchange service is a service that enables general calls through on-premise calls and public telephone networks occurring on the premises of a company or school. It is evolving from the public telephone network base such as network) to the Internet protocol (IP) network base.

Recently, the premises exchange service has provided various functions as well as a function of connecting a specific phone in the premises through a conventional key-phone function, that is, a representative number. Accordingly, each workplace has increased the desire to produce a variety of announcements according to the characteristics of the workplace, and to change the produced announcement as needed.

By the way, the announcement is generally implemented by recording the voice of the voice actor. That is, after deciding the sentences to be recorded in advance, the voice actors were recorded and voiced by the voice actors read the sentences, and the announcements were produced or changed in such a manner as to be applied to the system. Therefore, there is a problem that the production and change of the guidement takes a lot of time and money.

SUMMARY OF THE INVENTION The present invention has been made in an effort to provide an internal exchange system and a speech synthesis method.

In a method of synthesizing a voice for a announcement by a voice synthesis resource server in conjunction with a resource management server according to an embodiment of the present invention, the resource management server and the voice synthesis resource server are state information, wherein the state information is the voice. Exchanging an available channel of a synthetic resource server and server IP information of the available channel; When the available channel information is requested from the call box, the resource management server identifying an available channel of the voice synthesis resource server and transmitting server IP information of the available available channel to the call box; And receiving a speech synthesis request including sentences corresponding to the server IP information and announcement from the call box, wherein the speech synthesis resource server is a synthesizer engine, where the synthesizer engine stores the synthesized speech for a specific character or sentence. Synthesizing the speech using the synthesized speech DB; and generating a speech synthesis file for the sentence, and transmitting the result of the speech synthesis request to the call box.

An on-premises exchange apparatus according to an embodiment of the present invention includes a business system including a call box for serving a announcement when a call establishment request is received from an external user; A resource management server for identifying an available channel of a voice synthesis resource server and transmitting server IP information of the available available channel to the call box when receiving available channel information from the call box; And a voice synthesis request including sentences corresponding to the server IP information and announcements from the call box, a synthesizer engine, wherein the synthesizer engine uses a synthesized sound DB in which a synthesized voice for a specific character or sentence is stored. Synthesize a voice to generate a speech synthesis file for the sentence, and transmit the speech synthesis file to the call box.

According to an embodiment of the present invention, an on-premises exchange system and a speech synthesis method can be provided.

DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present invention. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and like reference numerals designate like parts throughout the specification.

Throughout the specification, when a part is said to "include" a certain component, it means that it can further include other components, without excluding other components unless specifically stated otherwise.

Now, a premises exchange system and a speech synthesis method according to an embodiment of the present invention will be described in detail with reference to the drawings.

1 is a diagram conceptually showing a premises exchange system according to an embodiment of the present invention. The premises exchange system 100 of FIG. 1 is constructed at one workplace and includes a voice synthesis server 102 for synthesizing a voice corresponding to the announcement at the request of the operator.

Referring to FIG. 1, the premises exchange system 100 according to an embodiment of the present invention includes a call box 101, a voice synthesis server 102, and a plurality of premises telephones 103_1, 103_2,... 103_n. .

The call box 101 is connected to the Internet network and performs a call setup procedure with one of a plurality of telephones 103_1, 103_2, ... 103_n based on a call setup request received through the Internet network. In addition, the call box 101 performs a call setup procedure between the plurality of internal telephones 103_1, 103_2, ... 103_n.

In addition, when the call box 101 receives a voice synthesis request from the operator to synthesize a voice corresponding to the announcement, the call box 101 requests a voice synthesis from the voice synthesis server 102 and then synthesizes the voice from the voice synthesis server 102. Receive and save the composite file. Then, when there is a call setup request from the Internet, the announcement service is played by playing the stored voice synthesis file.

The speech synthesis server 102 is connected to the call box 101, and when a speech synthesis request is received from an operator through the call box 101, synthesizes a speech corresponding to the announcement to generate a speech synthesis file, The generated speech synthesis file is transmitted to the call box 101.

The plurality of premises telephones 103_1, 103_2, ... 103_n refer to telephones installed in one business place. In this case, the telephones may include not only ordinary telephones but also faxes. In addition, the call box 101, the voice synthesis server 102, and the plurality of local telephones 103_1, 103_2, ... 103_n use xDSL (x digital subscriber line), which is a digital subscriber line that enables high-speed data communication. May be connected.

However, according to the premises exchange system of FIG. 1, since the voice synthesis server 102 must be constructed for each business location where the call box 101 is located, there is a problem that the construction cost is increased. Therefore, it is necessary to establish a server for speech synthesis separately from the workplace.

2 is a diagram conceptually showing an premises exchange system according to another embodiment of the present invention.

Referring to FIG. 2, the premises exchange system according to another embodiment of the present invention includes a voice synthesis resource server 200, a resource management server 300, and a workplace system 400.

The speech synthesis resource server 200 includes an initialization manager 201, a resource manager 202, a web manager 203, a timer manager 204, an engine manager 205, a plurality of synthesizer engines 206_1, 206_2,..., 206_n. ), And a plurality of DBs (207 to 212, 213_1, 213_2, ..., 213_m).

When the initialization manager 201 receives the initialization request from the resource manager 202, the initialization manager 201 initializes the plurality of synthesizer engines 206_1, 206_2,..., 206_n based on the information stored in the initialization DB 207. In addition, the initialization manager 201 manages initialization information that can be used when the engine manager 205 drives each of the plurality of synthesizer engines 206_1, 206_2,..., 206_n.

At this time, the initialization DB 207 is information for initializing the plurality of synthesizer engines 206_1, 206_2,..., 206_n, and the maximum storage directory and synthesizer engine used when the speech synthesis file is stored in the speech synthesis resource server 200. Number of log files (MaxChannel, n), number of log files (MaxLog), description level of log files (LogLevel), coding method for voice synthesis such as Vox, Wave (CodingFormat), voice actor ID (SpeakerID), voice Speed of synthesized file, Pitch of voice synthesized file, Volume of voice synthesized file, MaxResponseTime required for production when voice synthesized request (MaxResponseTime), CPU maximum utilization rate (MaxCPU), Maximum disk size Utilization (MaxDISK) and the like.

When the maximum utilization rate (CPU) of the CPU stored in the initialization DB 207 is exceeded, the speech synthesis resource server 200 cannot synthesize speech for a new announcement, and all of the channel states of the channel DB 208 are “busy”. Set to "."

The channel DB 208 stores state information of each channel. For example, if 30 is currently set as the maximum channel in the initialization DB 207, the channel DB 208 currently uses several channels out of all 30 channels, and some channels are idle. Information about whether or not the status is stored.

The resource manager 202 controls the overall operation of the speech synthesis resource server 200. In detail, the resource manager 202 requests the initialization manager 201 to perform initialization. In addition, the resource manager 202 regularly or irregularly transmits the contents stored in the channel DB 208 to the resource management server 300, so that the resource management server 300 is the channel state of the speech synthesis resource server 200. It can be identified in real time.

In addition, when the resource manager 202 receives a speech synthesis request from the call box 401, the resource manager 202 requests the speech synthesis from the engine manager 205 and requests the timer manager 204 to perform a timer.

Subsequently, when the voice synthesis completion response is not received from the engine manager 205 until the timer manager 204 receives the information indicating that the set time has elapsed, the resource manager 202 calls the call box 401 for requesting the speech synthesis. Send an incomplete message.

As a method of storing the generated speech synthesis file, a buffering method for storing the generated speech synthesis file in the call box 401 requesting speech synthesis and a filing for storing the generated speech synthesis file in the speech synthesis resource server 200 There is a way.

In the case of the buffering method, when the resource manager 202 receives the speech synthesis completion response from the engine manager 205, the resource manager 202 transmits the generated speech synthesis file stored in the resource DB 212 to the call box 401. do.

On the other hand, in the case of the filing method, when the resource manager 202 receives the speech synthesis completion response from the engine manager 205, the resource manager 202 checks the generated speech synthesis file stored in the resource DB 212 and calls the call box. Information about the completion of speech synthesis is transmitted to 401.

The web manager 203 allows a user to access and manage the operation DB 211 in which the operation information of the voice synthesis resource server 200 is stored through the web. In other words, the web manager 203 allows the user to access through the web without using a dedicated terminal to grasp and modify the contents of the operation DB 211.

When the timer manager 204 receives a timer request from the resource manager 202, the timer is activated, and when the time set in the timer DB 212 (for example, 3 seconds) is reached, the timer manager 204 indicates that the set time has elapsed. The information is passed to the resource manager 202.

The engine manager 205 manages generation and destruction of the plurality of synthesizer engines 206_1, 206_2,..., 206_n, and stores state information of the plurality of synthesizer engines 206_1, 206_2,..., 206_n in various DBs 213_1, 213_2. , ..., 213_m).

The plurality of synthesizer engines 206_1, 206_2,..., 206_n are based on the information stored in the initialization DB 207, and are generated in a manner of speech synthesis and are extinguished when the speech synthesis is completed. The plurality of synthesizer engines 206_1, 206_2,..., 206_n are capacity limits according to hardware performance, and are generated within the maximum number n of engines stored in the initialization DB 207. 208).

When the engine manager 205 receives a voice synthesis request from the resource manager 202, the information stored in the channel DB 208 and the initialization DB 207 indicates whether the synthesizer engines 206_1, 206_2, ..., 206_n can be generated. Judging from the base. When there is initialization data requested by the call box 401, the synthesizer engine 206_i is generated based on the initialization data received from the call box 401.

The plurality of synthesizer engines 206_1, 206_2,..., 206_n synthesize a voice for the guidement using a synthesized sound DB 213_1, 213_2,... 213_m that stores a synthesized voice for a specific character or sentence. Create a composite file. In this case, the generated speech synthesis file is reflected in the resource DB 210.

In the cache DB 209, the speech synthesis file generated by the engine manager 205 is buffered and stored in the cache DB 209 without running the synthesizer engine when requesting speech synthesis under the same condition from another call box. Speech synthesis file is used.

The resource DB 210 stores all the states after the call box 401 requests the voice synthesis resource server 200 for voice synthesis. Specifically, the request channel ID received from the call box 401 and the contents of the InfoDB 403 of the call box 401 are stored. In addition, the resource DB 210 stores information on whether the speech synthesis file and the speech synthesis result are normal, and timer information.

The resource management server 300 includes a server manager 301 and a web IF 302, and allocates resources of the speech synthesis resource server 200 to be used for speech synthesis at the request of the call box 401.

The server manager 301 receives the available channel information for speech synthesis from the call box 401, checks the available channel of the speech synthesis resource server 200 based on the contents stored in the server DB 303, and checks the call box ( 401) provides information about the server IP of the available channel.

In this case, the server DB 303 may store state information of the voice synthesis resource server 200, and specifically, the same information as that of the channel DB 208 of the voice synthesis resource server 200 may be stored.

The web IF 302 is managed in the same manner as the web manager 203 of the speech synthesis resource server 200 by accessing and managing the operation DB 304 in which the operation information of the speech synthesis resource server 200 is stored through the web. Do it. At this time, the operation DB 304 includes various DB (207 to 212, 213_1, 213-2, ..., 213_n) information in the speech synthesis resource server 200.

When the contents of the operation DB 304 are modified by the operator, the web IF 302 transmits the modified contents to the speech synthesis resource server 200, and the speech synthesis resource server 200 receiving the modified contents is The contents are reflected in various DBs 207 to 212, 213_1, 213-2, ..., 213_n in the voice synthesis resource server 200. Through this, the speech synthesis resource server 200 and the resource management server 300 will always have the same information.

The workplace system 400 includes a call box 401, a plurality of DBs 402-405, and a plurality of campus telephones 103_1, 103_2,... 103_n.

The call box 401 is connected to the Internet network and performs a call setup procedure with one of the plurality of premises telephones 103_1, 103_2,... 103_n based on the call setup request received through the Internet network. In addition, the call box 401 performs a call setup procedure between the plurality of internal telephones 103_1, 103_2, ... 103_n.

In addition, when the call box 401 receives a voice synthesis request from the operator, the call box 401 requests channel information available to the resource management server 300 and receives server IP information of available channels from the resource management server 300. Thereafter, the call box 401 transmits the sentence corresponding to the server IP information of the available channel, the Info DB 403, and the announcement to the speech synthesis resource server 200 and requests speech synthesis.

The speech synthesis request from the operator may be specifically implemented in the following manner. When the operator receives the information on the speech synthesis request from the operator, the call box 401 displays a speech synthesis request screen for the announcement and requests the user to input the contents of each area on the screen.

Subsequently, when the buffering method is used, the voice synthesis file is received from the voice synthesis resource server 200 and stored in the TTS Wave DB 402. When a call setup request is received from the Internet, the TTS Wave DB 402 is received. The announcement is serviced by playing the stored speech synthesis file.

In this case, when the operator receives the information on the speech synthesis request, the call box 401 is displayed to the user and the voice synthesis request screen requesting to input the contents of each area may be implemented as shown in FIG. 3.

3 is a diagram illustrating a speech synthesis request screen according to an embodiment of the present invention. Referring to FIG. 3, the voice synthesis request screen 30 includes a voice title area 31, a voice file name area 32, a voice file location area 33, a voice production text area 34, and a voice file format area ( 35, a voice actor selection area 36, a comment area 39, and a producer name area 40, and a voice production performance icon 37 and a pre-listening icon 38.

The operator inputs a title for the announcement in the voice title area 31 and inputs a name to be used for the voice synthesis file to be generated in the voice file name area 32. In addition, the voice file location area 33 inputs the location where the voice synthesis file to be stored is stored and inputs the text corresponding to the announcement to generate the voice synthesis file in the voice production text area 34.

 In addition, a coding scheme (Vox, Wav, ...) of a speech synthesis file to be generated is input to the voice file format region 35, and a voice actor voice to be synthesized is input to the voice actor selection region 36.

After inputting each of the above areas 31 to 36, the operator clicks the voice production performing icon 37 to request the voice synthesis from the call box 401. In this case, the call box 401 transmits the contents input to each area 31 to 36, the Info DB 304, and the server IP information received from the resource management server 300 to the voice synthesis resource server 200. And request voice synthesis.

Subsequently, when the buffering method is used, the call box 401 which receives the speech synthesis file for the announcement from the speech synthesis resource server 200 stores the received speech synthesis file in the TTS Wave DB 402.

Then, when the operator clicks on the pre-listen icon 38, the call box 401 plays the speech synthesis file stored in the TTS Wave DB 402.

When the operator listens to the reproduced voice and the operator has a comment on the complementary requirements for the synthesized file, the operator inputs a comment in the comment area 39 of the voice synthesis request screen 30, and the inputted comment is a comment DB. 404 is stored as comment content for the speech synthesis file. In addition, the operator may enter his name in the producer name area (40).

Referring back to FIG. 2, the TTS Wave DB 402 stores a speech synthesis file for the announcement synthesized by the speech synthesis resource server 200.

The Info DB 403 stores communication protocol information for communication between the call box 401 and the voice synthesis resource server 200, and specifically, a delimiter (BufferORFile) for distinguishing a buffering method or a filing method (BufferORFile) and a sentence length (MessageLength). ), Voice synthesis processing result (nResult), coding method (nFormat) when generating a voice synthesis file, received server IP (ServerIP), synthesized sound DB information (SpeakerID) required when generating a voice synthesis file, volume information (Volume), speed Information (Speed), pitch information (Pitch), and length information (TTSLength) of the speech synthesis file are stored.

In the comment DB 404, the operator's opinion about the speech synthesis file generated by the speech synthesis resource server 200, specifically, the operator inputs in the comment area 69 of the speech synthesis request screen 60 of FIG. The contents are saved. The contents stored in the comment DB 404 may be used later in the voice synthesis resource server 200 for tuning the synthesized sound DBs 213_1, 213_2,... 213_m.

The scenario DB 405 stores a format used when an operator requests speech synthesis. In this case, the format used may specifically be the screen format of FIG. 3.

Now, a speech synthesis method according to an embodiment of the present invention will be described in detail with reference to the drawings.

4 is a flowchart illustrating a speech synthesis method according to an embodiment of the present invention.

Referring to FIG. 4, the resource management server 300 and the voice synthesis resource server 200 reflect state information between each other (S401). The speech synthesis resource server 200 transmits the number of channels available during operation to the resource management server 200 regularly or irregularly, and the resource management server 200 changes when the information of the operation DB 304 is changed. The contents are transmitted to the speech synthesis resource server 200 to reflect state information with each other.

For speech synthesis, the call box 401 first requests available channel information of the speech synthesis resource server 200 to the resource management server 300 (S402), and the resource management server 300 is stored in the operation DB 304. The available channel is identified based on the information, and the server IP information of the available channel is transmitted to the call box 401 (S403).

The call box 401 receiving the server IP information transmits the sentences corresponding to the received server IP information, the Info DB 403, and the announcement to the speech synthesis resource server 200 and requests speech synthesis (S404). .

The speech synthesis resource server 200, which has received the speech synthesis request, produces a speech synthesis file based on the received information (S405), and transmits the generated speech synthesis file to the call box 401 by the buffering method. (S406). The call box 401 stores the received speech synthesis file in the TTS Wave DB 402, and when there is a call setup request from the Internet network, plays back the speech synthesis file stored in the TTS Wave DB 402. Service the comment.

In the above, the case of storing the speech synthesis file in a buffering manner has been described. However, the buffering method has a problem that an error occurs due to data transmission, and there are two solutions to this problem.

In the first method, the voice synthesis resource server 200 has the same data structure of the Info DB 403 of the call box 401. The second method first receives information based on the information of the Info DB 403. There is a method of receiving after confirming the amount of information.

Referring to the second method in detail, when the call box 401 requests the speech synthesis to the speech synthesis resource server 200, the call box 401 first transmits the Info DB 403, and later transmits a sentence corresponding to the announcement. That is, the speech synthesis resource server 200 checks the sentence length of the announcement to be received based on the information of the MessageLength in the Info DB 403, and then enters the announcement in the reception mode until the sentence length stored in the MessageLength comes. Receive the corresponding sentence.

Then, when all data corresponding to the sentence length stored in the MessageLength from the call box 401, the speech synthesis resource server 200 synthesizes the voice for the received announcement.

In addition, after the speech synthesis file is generated for the announcement, the speech synthesis resource server 200 determines the size of the speech synthesis file stored in the cache DB 209 and stores the size in TTSLength of the Info DB 403. Input and send to the call box (401).

Thereafter, the call box 401 checks the value stored in the TTSLength of the received Info DB 403, and then receives the speech synthesis file in the reception mode by the stored size. Therefore, according to the second method, it is possible to reduce data errors due to buffering.

In the above, only the method of storing the speech synthesis file in the buffering method has been described. However, when the BufferORFile of the Info DB 403 is used, the buffering method and the filing method may be simultaneously implemented.

5 is a flowchart in which a buffering scheme and a filing scheme are simultaneously implemented according to an embodiment of the present invention.

Referring to FIG. 5, the call box 401 allocates one of a buffering method or a filing method as information about a method to be performed in the BufferORFile of the InfoDB 403 (S501), and receives the server from the resource management server 300. The IP information, the InfoDB 403, and the sentence corresponding to the announcement are transmitted to the speech synthesis resource server 200, and the speech synthesis is requested (S502).

The speech synthesis resource server 200 that has received the speech synthesis request determines whether the buffering scheme or the filing scheme is implemented based on the value stored in the BufferORFile of the InfoDB 403 (S503).

When implemented in a buffered manner, the speech synthesis resource server 200 generates a speech synthesis file for the announcement and uses the speech synthesis file as the cache DB 209 (S504 and S505). Thereafter, the size of the speech synthesis file is determined, and the size determined in the TTSLength of the InfoDB 403 is allocated (S506 and S507).

Thereafter, the speech synthesis resource server 200 transmits the InfoDB 403 to the call box 401, and then transmits the generated speech synthesis file to the call box 401 (S508 and S509). After the speech synthesis file is transmitted to the call box 401, the synthesis engine that synthesizes the speech for the announcement in the speech synthesis resource server 200 is extinguished (S510).

When implemented by the filing method, the speech synthesis resource server 200 generates a speech synthesis file for the announcement and uses the speech synthesis file as the cache DB 209 (S511 and S512). Thereafter, the speech synthesis file is stored in the location stored in the SavedDirectory of the InfoDB 403, and the InfoDB 403 is transmitted to the call box 401 (S513 and S514). Thereafter, the synthesis engine, which synthesized the voice for the announcement in the speech synthesis resource server 200, is extinguished (S515).

The embodiments of the present invention described above are not only implemented by the apparatus and method but may be implemented through a program for realizing the function corresponding to the configuration of the embodiment of the present invention or a recording medium on which the program is recorded, The embodiments can be easily implemented by those skilled in the art from the description of the embodiments described above.

Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements of those skilled in the art using the basic concepts of the present invention defined in the following claims are also provided. It belongs to the scope of rights.

1 is a diagram conceptually showing a premises exchange system according to an embodiment of the present invention.

2 is a diagram conceptually showing an premises exchange system according to another embodiment of the present invention.

3 is a diagram illustrating a speech synthesis request screen according to an embodiment of the present invention.

4 is a flowchart illustrating a speech synthesis method according to an embodiment of the present invention.

5 is a flowchart in which a buffering scheme and a filing scheme are simultaneously implemented according to an embodiment of the present invention.

Claims (12)

In the speech synthesis resource server in conjunction with the resource management server to synthesize the speech for the announcement, Exchanging status information between the resource management server and the speech synthesis resource server, wherein the status information includes an available channel of the speech synthesis resource server and server IP information of the available channel; When the available channel information is requested from the call box, the resource management server identifying an available channel of the voice synthesis resource server and transmitting server IP information of the available available channel to the call box; And When receiving a speech synthesis request including sentences corresponding to the server IP information and announcement from the call box, the speech synthesis resource server is a synthesizer engine, where the synthesizer engine stores the synthesized speech for a specific character or sentence. Synthesizing the voice using the synthesized sound DB; and generating a voice synthesis file for the sentence, and transmitting the result of the voice synthesis request to the call box. Speech synthesis method. The method of claim 1, The result of the speech synthesis request is the speech synthesis file, Storing, by the call box, the speech synthesis file, the speech synthesis file; And If the call box receives a call establishment request, further comprising reproducing the speech synthesis file to service announcements; Speech synthesis method. The method of claim 1, Exchanging the status information, Transmitting, by the voice synthesis resource server, information on an available channel to the resource management server; And If the information stored in the resource management server is changed, transmitting the change to the speech synthesis resource server. Speech synthesis method. The method of claim 1, Sending the result of the speech synthesis request to the call box, Receiving, by the speech synthesis resource server, information on the total length of the sentence from the call box; And When the speech synthesis resource server operates in a reception mode until the sentence length received by the information on the entire length of the sentence is received, and receives all data corresponding to the entire length of the sentence, the voice for the data Comprising the steps of synthesizing Speech synthesis method. The method of claim 1, Sending the result of the speech synthesis request to the call box, Determining, by the speech synthesis resource server, the length of the speech synthesis file and transmitting the determined length of the speech synthesis file to the call box: And Operating the call box in a receive mode until a length of the speech synthesis file is received. Speech synthesis method. The method of claim 1, Sending the result of the speech synthesis request to the call box, Information about a method of storing the speech synthesis file from the call box by the speech synthesis resource server, a buffering method of storing the speech synthesis file in the call box, and the speech synthesis file stored in the speech synthesis resource server. Receiving one of the filing schemes Speech synthesis method. The method of claim 1, Sending the result of the speech synthesis request to the call box, Generating, by the speech synthesis resource server, a synthesizer engine to generate a speech synthesis file for the sentence, and extinguishing the synthesizer engine after the speech synthesis file is generated. Speech synthesis method. The method of claim 1, Sending the result of the speech synthesis request to the call box, When the speech synthesis resource server receives a filing scheme from the call box as information on the manner of storing the speech synthesis file, storing the speech synthesis file at a location preset in the speech synthesis resource server. More containing Speech synthesis method. A business system including a call box for providing announcements when receiving a call setup request from an external user; A resource management server for identifying an available channel of a voice synthesis resource server and transmitting server IP information of the available available channel to the call box when receiving available channel information from the call box; And When receiving a voice synthesis request including a sentence corresponding to the server IP information, announcements from the call box, a synthesizer engine, where the synthesizer engine uses a synthesized sound DB that stores a synthesized voice for a specific character or sentence. A speech synthesis resource server for generating a speech synthesis file for the sentence by synthesizing speech and transmitting the speech synthesis file to the call box. Premises exchange unit. 10. The method of claim 9, The workplace system, Further comprising a TTS Wave DB for storing the speech synthesis file Premises exchange unit. 10. The method of claim 9, The speech synthesis resource server A resource manager for transmitting information about an available channel to the resource management server; And And a engine manager for generating a synthesizer engine to generate a speech synthesis file for the sentence, and destroying the synthesizer engine after the speech synthesis file is generated. Premises exchange unit. 10. The method of claim 9, The resource management server A server manager for receiving available channel information for voice synthesis from the call box and providing server IP information of the available channel to the call box; And If the information stored in the resource management server is changed, WebIF for transmitting the change to the voice synthesis resource server Premises exchange unit.

Images