TW424373B - Device for speech realization of caller's incoming message and process method thereof - Google Patents

Abstract

The present invention relates to a device for speech realization of caller's incoming message and the process method thereof. The present device uses voice to express the caller ID and the text of the caller's short message. The present device utilizes a phone book database for recording the caller's name and its corresponding caller ID and a voice database for recording the phonemic units and the corresponding phonemic data. When receiving an incoming caller ID, the device can search through the phone book database to locate the corresponding caller's name before searching through the voice database to locate the phonemic data corresponding to the caller's name. When receiving an external short message, the device can search through the voice database directly to locate the phonemic data corresponding to the short message. After synthesizing the phonemic data by a digital signal processor to generate voice signals, the message can be played, thereby accomplishing the purpose of speech realization.

Description

424373 V. Description of the invention (l) ----- The present invention relates to a communication service, in particular, to provide a caller's incoming message voice implementation device and processing method, which can target mobile communication such as mobile phone The device uses voice mode to call the incoming caller's message to achieve a better man-machine interface. With the increasing popularity of mobile phones and the rapid development of mobile communication services, more and more customers have begun to use some additional communication services, such as call indication function and short message service (SMS) ^ Most telecom service providers also provide these services to attract customers. The call indication function is to use the headend device or the switch to process the call exchange, and the caller's phone number (caller-ID) can be transmitted to the receiver's mobile phone. The receiver's mobile phone is receiving When the caller's phone number is reached, it will be displayed on the liquid crystal (LCD) screen of the mobile phone, so that the customer can know the caller's phone number before receiving the call, and the short message service is the caller's use of short text paragraphs To express the message to be transmitted, the receiver's mobile phone can use the LCD screen to display this short text paragraph. The above services are mainly achieved using digital mobile communication networks, that is, digital data is used to transmit the caller's phone number in the call indication function and the text of the short message in the short message service. But unfortunately, the current call indication function and short message service on mobile phones can only pass messages through the ringing of the phone and the LCD screen. Such a monotonous human-machine interface sometimes does not provide customers with the most convenient services. After all, mobile phones are a voice communication service. If you can

^ 2 ^ 3 ~ 7 3 1________ V. Description of the invention (2) Voice of various functions of the mobile phone system as much as possible, which is helpful to improve the service quality of the mobile phone. In view of this, the main purpose of the present invention is to provide a caller's incoming message voice implementation device and processing method, so that the caller's phone number and text message can only be communicated by the LCD screen in the past, and can be directly transmitted through voice Way to achieve, increase the service quality of mobile phones. According to the above object, the present invention proposes a caller's incoming message voice implementation device, which is suitable for use in a mobile communication device, such as a mobile phone, to communicate the caller's phone number and short message text by voice. This device includes at least two memories. One is used to store the first memory of the phone book database. Since the address book database can be edited by customers, it is best to use the erasable programmable program. Read e-memory (EPR 0 M) or flash memory to implement. The address book database is composed of communication items, each communication item records the caller name and its corresponding caller ID (caller ID); the other is used to store the second memory of the voice database. Some data does not need to be edited by the customer, so read-only memory (ROM) can be used to reduce production to f. The speech database is composed of text unit items, and each text unit item is used to record phoneme units and their corresponding phoneme data. In addition, 'this voice database can be a database created by using text-to-speech (Text τ〇, TTS for short) technology, or a database created by using real-life audio technology, and a voice database created by using TTS technology. It is relatively small, so it is also more suitable for mobile phone applications. Page 5 ^ 24373 The voice of the device number and the calling party name of the control unit is broadcasted through the second data, and it can be used for ringing or tone. Unit, processing. A memory is transmitted through the mobile phone real-time data processing sheet received by the phoneme asset generation unit. Therefore, there are five descriptions of the invention. (3) In addition, the caller's telephone code number is controlled to query the corresponding database and query the call manager. The composition of the broadcasting unit can pass through the corresponding phoneme number, and then re-speech the text. The service quality of using the complete language in the phone also includes the control of short messages. The third party name can be used, and then the corresponding processing unit is called. For speech from external memory, speech sounds are reused. In addition, the same method is used to achieve the liquid crystal display mode, which is used for the address book and the second memory of the incoming call. Use the short-message database after the digital signal to query the meta-synthesis generated by the phone. Due to the delivery of the message, the received caller's electrical database is effectively enhanced, and the voice data bit signal is 'reused' to control the incoming call of the voice message in the single text message. The mobile phone can now be operated. In addition, the present invention A method for processing the caller's incoming call voice is also provided. In this method, a contact database and a voice database are first provided. The contact database can be used to record the caller's name and its correspondence. The caller's phone number, and the voice database is used to record phoneme units and their corresponding phoneme data. When the above mobile phone receives an incoming call, 'the caller's phone number is retrieved. Then, the address book database is used to query the corresponding caller name, and then the voice database is used to query the corresponding caller name. The included phoneme data β finally synthesizes the obtained phoneme data to generate a sound signal for playback, so the customer can directly hear the name of the caller who played the voice. For SMS

4243 7 3 V. Description of the invention (4) The short message in the messaging service and the preset incoming voice text can also be directly queried through the voice database to query the corresponding phoneme data, and then synthesize it to produce a sound signal and play it . Brief description of the drawings: In order to make the above-mentioned objects, features, and advantages of the present invention more comprehensible, a detailed description is given below with a preferred embodiment 'in conjunction with the attached drawings' as follows: FIG. 1 shows the present invention Schematic diagram of setting the incoming call voice on the mobile phone to achieve the related settings in the embodiment.

Fig. 2 shows a block diagram of a mobile phone having a caller's incoming message voice implementation device D in the embodiment of the present invention. Fig. 3 shows a flowchart of the caller's incoming call indication voice implementation function in the embodiment of the present invention. FIG. 4 shows a flowchart of the short message service voice implementation function in the embodiment of the present invention. Explanation of symbols: 1 ~ LCD display screen; 3 ~ Antenna; 5 ~ Keyboard: 7 ~ Playback device; 9 ~ Radio receiver; 10 ~ Microprocessor; 12 ~ RF circuit; 14 ~ Digital signal processor (DSP); 1 6 ~ Random access memory (KAM); 1 flash memory; 20 read-only memory (ROM); 28 ~ address book database; 30 ~ TTS database. Embodiment: The device and method for realizing the incoming call voice of the present invention mainly propose an improved solution for some communication services of a general mobile phone, that is, adding a voice prompt function to make the function more perfect. The following cooperation diagram

Page 7 4 ^ 24373 ___ V. Description of the invention (5) Formula ′ will be described in detail in an embodiment. The mobile phone communication services used in this embodiment for description include three types, which are caller's incoming call indication, short message service, and ringing prompt. In the caller's incoming call indication function, this embodiment can convert the caller's phone number (caller-ID) sent by the switch to the real name of the caller, and then send it out using simulated human voice. In the short message service, this implementation For example, the caller's message can be processed in the above manner, and for the short message to be displayed, it can be directly issued using simulated human voice. In the vibrating alert function, this embodiment is a preset reminder text. At the time, the voice of the preset prompt text is generated by using a simulated human voice to replace the traditional ringing prompt. However, a general mobile phone user does not necessarily need such a voice prompt function at any time, so in this embodiment, a setting method is also provided for the user to choose whether to use this voice function. Fig. 1 shows a schematic circle of "setting an incoming message voice on a mobile phone to realize related settings in this embodiment." As shown in the figure, on the LCD (liquid crystal display) 1 of the mobile phone, three setting options are listed, which are "Incoming Call Instruction Voice Implementation", "Short Message Service Voice Implementation", and π Vibration Prompt Voice Implementation. ". The user can adjust the selection of three settings through the selection key and the setting key on the keyboard 5. The setting status of these three voices is “Switch on” or “Close π”, which is to choose whether to enable this function. After the setting is completed, the setting values of these states are stored in the internal memory device (not (Illustration). Except where noted below, all three states are set to "Enable" for explanation. Figure 2 shows the implementation of the caller's incoming call voice in this embodiment. Page 8

V. Description of the Invention (6) Block diagram of the device's mobile phone. In Figure 2, the mobile phone includes 1, antenna 3, keyboard 5, broadcast device 7, radio device 9, micro-processing, 10, radio frequency circuit 12, DSP (digital signal processor) 14, and memory (including RAM 1 6 'Flash memory 18, ROM 20). The LCD 1 is used as the output screen of the mobile phone to display related information; the keyboard 5 is used as the data input device of the mobile phone, allowing the user to input instructions or related data, such as a person's name or phone number. The receiving or playing of the sound part is handled through the radio device 9 and the broadcasting device 7, such as a general micro microphone and a micro speaker. The antenna 3 is coupled to the radio frequency circuit 12 for receiving or transmitting information from the remote switch to the remote switch. The radio frequency circuit 12 is used for processing a part related to a radio frequency (RF) signal. DSP 1 4 is used to perform digital data related processing. For example, using TC 9 3 3 1 'and microprocessor 1 〇 as the control center of mobile phones can use MI PS R 2 0 0 〇 or other similar Microprocessor. It must be noted that although in some mobile phone designs, Dsp 14 and microprocessor 10 are implemented on the same chip, they do not depart from the spirit of implementation of the present invention. In addition, a variety of different memory types are suitable for storing data with different needs, such as RAM 16, which cannot retain the data stored in it after the power is turned off. Therefore, it is used to store various temporary data; ROM 2 It stores fixed data that cannot be changed in advance. It is suitable for saving system execution code or fixed database. Flash memory 18 is in between, which is erasable and programmable. 1. read-only memory

Pr0gra (able R0M) type, its internal data can be changed 'while the internal data can be maintained after the power is turned off-a considerable period of time, because of 424373 V. Description of the invention (7) This can be used to store user-set Information, such as contacts. This embodiment is related to the microprocessor 10, the DSP 14, the broadcasting device 7, the flash memory 18, and ROM 20, and the functions of these components in this embodiment will be described in detail below. In this embodiment, two databases must be used, namely the address book database 28 and the TTS (Text To Speech ') database 30, which are stored in the flash memory 18 and ROM 20, respectively. The address book database 28 is data created and edited by the user 'and is therefore stored in the flash memory 18 which can be changed. The address book database 28 is composed of several communication items. Each communication item records the caller's name (ie, the name of the person) and its corresponding caller's phone number (ie, its phone number). Its function is that when the mobile phone receives an incoming call and retrieves the caller's phone number, it can use the content in the address book database 28 to find the corresponding person name. Since a general mobile phone has a function of designing an address book originally, using this function to establish the address book database 28 in this embodiment can achieve the purpose of sharing. TTS database 30 is a speech database created using TTS phoneme speech synthesis technology. The basic working principle of TTS phoneme speech synthesis technology is to use different pronunciation units (also called phonemes) to spell out the pronunciation of a word. For a Chinese word, 'corresponds to the initial and final phonemes; for an English word,' corresponds to the phonemes of vowels and consonants. For detailed technical contents, please refer to the Republic of China Patent Case Nos. 851 1 2444 and 851 1 2445. In the TTS database 30 of this embodiment, it is composed of a plurality of text unit items. Each of the text unit items records a phoneme unit and its corresponding phoneme data. When synthesizing text into speech, 'According to known pronunciation rules, 424373 can be used. 5. Description of the invention ⑻ The text is first decomposed into phoneme units, and then the corresponding phoneme data is searched through the database! 30. Finally, it can be They are synthesized using a digital signal processor. In addition, the speech synthesis method of the present invention can also be implemented by using human voice technology. Real-life vocalization technology synthesizes speech through a large-scale real-life vocal database. However, in this embodiment, the TTs phoneme speech synthesis technology is still taken as an example, mainly because it can already achieve a fairly realistic analog human voice, and at the same time, the cost is low. Taking Chinese as an example, all Chinese phoneme units do not exceed approximately 1,500 ’and the capacity of the TTS phoneme database does not exceed 1 MB. Compared to the large amount of hardware resources required for real-life voice technology, TTS technology is more suitable for light and thin products such as mobile phones. In this embodiment, the microprocessor 10 is used to control the execution flow of the voice implementation. That is, when the voice implementation function is activated, the microprocessor J 0 uses the address book database 28 and the TTS database 30 to convert the message into The corresponding phoneme data is then synthesized by the DSP 14 into a sound signal, which is delivered to the broadcasting device 7 for broadcasting. Since the signal demodulation part of the baseband 1C in a general mobile phone includes a digital signal processor, the DSP 1 4 can also be implemented using this original component to reduce costs. As described above, "this embodiment is processing for implementing voice services for three services" The following describes the processing methods for implementing voice services for these three services in detail based on the above-mentioned system architecture. Calling Party Instruction Voice Realization Function Figure 3 shows a flowchart of the calling party's incoming voice instruction realization function in this embodiment. First, when the caller calls, the program of the caller's call indication voice in the microprocessor 10 is started (S1). Follow up ΐΗΠϋΗ " 1 — — — — Page 11 4M3 73 «V. Description of the Invention (9) The phone will demodulate the modulation message sent by the switchboard and retrieve the caller's phone number (caller_ID) for this incoming call. ), Which is the phone number of the caller (S2). The microprocessor 10 then checks whether the voice realization function is enabled (S3). If it is set to the off state, it processes the incoming call according to the general mode (S4), such as displaying the phone number of the incoming call directly on the LCD 1; If set to the enabled state, the following voice realization processing is performed. At this point, the microprocessor 10 has obtained the telephone number of the caller. Therefore, the microprocessor 10 finds the caller's name through the address book database 28 stored in the flash memory 18 based on the caller's telephone number (S5). At this time, the caller's name is a text type, so the microprocessor 10 cuts the found caller's name into phoneme units according to the established rules of TTS speech synthesis, and then uses the TTS database 30 stored in ROM 20 to find the corresponding All phoneme data of (S6). After the search of the phoneme data is completed, the microprocessor 10 sends the phoneme data to the DSP 14 for TTS phoneme synthesis to generate a sound signal corresponding to the caller's name (S7), and finally uses the broadcasting device to send out this sound signal (S8). This completes the procedure for voice instructions for incoming calls. Users can use the normal call handling method to handle subsequent actions. In addition, in this embodiment, when an incoming call occurs, in addition to using the voice to perform an incoming call indication as described above, the incoming call number can also be displayed on the LCD 1 at the same time. Example 1 This example illustrates the actual operation of the caller's call indication voice implementation function. Assume that a communication item stored in the user's mobile phone address book database is (Yuan Ye 021-62539088). When Mr. Yuan Ye also called

Page 12 412 4373 V. Description of the invention (ίο) When this user's mobile phone is used, the telephone number of Mr. Yuan Ye will also be transmitted to the user's mobile phone via radio signal after being modulated ^ After being demodulated After that, the user's mobile phone can obtain the caller's phone number of 022, 62539088. Then according to the stored address book database 28 ', it can be found that this caller is "Yuan Ye". The microprocessor 10 "texts" Yuan Ye "to disassemble the phoneme unit" and then uses the TTS database 30 Find out all the corresponding phoneme data 'in conjunction with a fixed instruction text' Mr. Call '& the corresponding phoneme data' is sent to the DSP 14 for speech synthesis and playback. At the end, the mobile phone will issue a voice of "Mr. Yuan Ye calls" to indicate the caller's incoming call, instead of the traditional ringing instruction. Short Message Service Voice Implementation Function Figure 4 shows a flowchart of the short message service voice implementation function in this embodiment. The voice function of this short message service is mainly to play the name of the caller and the text of the short message transmitted by voice. In this function, since the user does not necessarily want the short message content itself to be played by voice, a setting value is added in this embodiment to allow the user to decide whether it is a confidential state, that is, whether to play the short message content β, and then according to Figure 4 illustrates the operation flow of this function. First of all, when the calling party calls to send a short message, the short message service voice realization program in the microprocessor 10 is started (sl0). Then the mobile phone will demodulate the modulation message sent by the exchange. This includes the caller's phone number (caller ID), which is the caller's phone number 'and the text of the short message to be transmitted (S11). Then the microprocessor 10 will check whether this voice realization function is enabled (S12). If it is set to OFF Τ1ΒΠ1 " "% 13 1 ^ 143 73 ^ _ _ V. Description of the invention (11) Closed state, according to the general Mode to process this incoming call (S1 3), such as displaying the phone number and short message of the incoming call directly on the LCD 1; if it is set to the 'start state', the following voice implementation processing is performed. Then, the microprocessor 10 first finds the caller's name through the address book database 28 stored in the flash memory 18 according to the caller's phone number (S14). Then according to the established rule of TTS speech synthesis, the found caller name is cut into phoneme units 'and then the TTS database 30' stored in ROM 20 is used to find all corresponding phoneme data (S15). The next step is to deal with the part about the text of the short message. The microprocessor 10 will first check whether the user is set to the secret mode (S16) ', that is, the content of the short message is not expected to be played by voice. If it is in the confidential mode, then the subsequent short message voice action is not performed (S1 7, S1 8), and skip to the speech synthesis step (S19) directly. »If it is not in the confidential mode, the microprocessor 10 will first obtain The content of the short message (S1 7), and then using the established rule of TTS phoneme speech synthesis technology, this short message content is decomposed into several phoneme units, and then the corresponding phoneme data is found using the TTS database 30 (S18). Whether it is the caller name phoneme data obtained from step S15 or the short message content phoneme data obtained from step S18, the DSP 14 performs TTS phoneme synthesis to generate a corresponding sound signal (S19), and finally uses the broadcasting device to send out This sound signal (S20). When in the privacy mode, ‘only the caller ’s incoming call-heart is played, but when in the privacy mode’, it includes the caller ’s name and the text message. In addition, in the short message service of this embodiment, in addition to using the voice for short message playback as described above, the short message text can also be

A 42437g Fifth, the invention description (〗 2) is displayed on the LCD 1. Example 2 This example illustrates the actual operation of the short message service voice implementation function. Continuing the situation in Example 1, when Mr. Yuan Ye wants to send a message "Merry Christmas" to the user's mobile phone, Mr. Yuan Ye's electric number and text content of the short message will be transmitted via radio signal after being adjusted. ^ To the user's mobile phone. After demodulation processing, use ^ to obtain the phone number of the caller 02 0262539088. According to the address book database 28 stored therein, you can find out which caller is " Yuan also ". Then the text "Yuan Ye" disassembles the phoneme unit, and then uses the TTS database 30 to find all the corresponding phoneme data. If in the non-secret mode at this time, the microprocessor 10 will similarly disassemble the phoneme unit of the short message "Merry Christmas", and then use the TTS database 30 to find all corresponding phoneme data "Finally, cooperate with __ fixed The phoneme information corresponding to the instruction text "Mr. Call" is sent to Dsp 14 for speech synthesis and the voice "Mr. Yuan Ye calls Merry Christmas" will be played, indicating the caller's incoming message β vibration钤 Prompt for voice realization function 钤 Indication of voice realization function is simpler than the previous two functions. It mainly uses a piece of preset text to replace the traditional ringing ringtone. Although many mobile phones on the market generally have The function of setting different snoring sounds, but using voice instructions is obviously better. When a mobile phone receives an incoming call, the microprocessor 10 takes out a preset voice message of an incoming call, such as "Someone calls you, Michael" * will After disassembling the corresponding phoneme unit, through the TTS database on ROM 20, query 424373. V. Description of the invention (13) The phoneme data corresponding to this caller's voice text. Similarly, these phoneme data are sent to DSP 14 for synthesis and broadcasting. In this function, because the content is fixed and the address book database is not needed, it is also feasible to use real-life pronunciation technology to implement it. Although the present invention is disclosed above with a nephew example, the position of Liu Beiqi is used to limit the present invention. Anyone skilled in this art can make some changes and retouch without departing from the core of the hair and scope, because two: The scope shall be determined by the scope of the attached patents. # 月 之 保护

Page 16

Claims (1)

424373 Six types of patent application scope Caller's incoming message voice implementation device, which can be placed in a line of communication equipment j, which includes: ~ 1, This system is used to store an address book database, the above address book-the exchange Make up 'Each of the above $$ items is used to call the party's name and its corresponding caller's phone number ^^ »-# _ °° __]] .__ stem_, used to store-voice database, the above voice data The library is composed of 77 ^ 1 ^ phoneme units and their corresponding phoneme data; ~, which is coupled to the above first memory and the above second note # which is called according to the above-mentioned mobile communication device. I—editor Through the above! A JHilf directory database, query the corresponding caller name, and according to the name of the query, through the above-mentioned ϋ database, query the corresponding Γ-number d-ί of the caller name ~~-The voice processing $ element is consumed by the above-mentioned control unit to receive the above-mentioned first phoneme data from the query and synthesize to generate an H · I »-I- .1. ^ Ιι.ι. ι ^, and — coupled to the above speech unit ’Is used to broadcast the first audio signal. '~ --- --------------- 2 · The caller's caller's incoming message voice as described in item f of the patent scope], where the above ^^ unit is further received by the above The mobile communication device receives one from outside ^ 2, and searches for the plural corresponding to the above information through the above-mentioned match 1-2 g 'ϊϊ ^ χχ · ^; the phoneme data management unit is ττ Said the second phoneme data, together with m One second Page 17 6'Patent Application Scope Sound Signal 'and broadcast through the above-mentioned broadcast unit. 3. The calling party's incoming call voice implementation device as described in item 1 of the scope of patent application, wherein when the mobile communication device receives an incoming call, the control unit takes out a preset word and uses the third memory. In the above-mentioned speech database, check Λ'ΐ'ϊ'ϊ 'for the plural number corresponding to the electric speech text. The above-mentioned speech processing unit receives the above-mentioned third tone data, and generates a third The sound signal 'is broadcast through the above-mentioned element '51 ¾. ------------ ^ erW · 4. The caller's incoming call voice implementation device described in item 1 or 2 or 3 of the scope of patent application, which further includes a liquid crystal display screen coupled to The above color system element is used when the above-mentioned caller telephone number of the caller, 'through the above-mentioned address book data bank, query to find a corresponding one, then the name of the caller is output in the above solution. Fold. ----— The calling party's incoming call funnel described in the item is erasable and programmable 5. If the patent application scope is 1 or 2 or 3, the voice-realizing device is the first read-only memory (EPKOM) . · ~~ 6 · The device for realizing the call voice according to item 1 or 2 or 3 of the scope of patent application, wherein the above-mentioned second party calls (ROM). '-Is a read-only memory telecommunication to a telecommunication telecommunication * The call-to-speech implementation device described in item 1 or 2 or 3 of the scope of patent application, wherein the above-mentioned voice information ^ ^ ^ (Text To Speech)' 8 * 如Call for patents described in item 1 or 2 or 3 6. Scope of patent application The above-mentioned voice database is a database established by using real human voice. The above 9. The caller's incoming message voice implementation device as described in the scope of patent application No. 丨 or 2 or 3, which makes the above-mentioned voice processing unit a bit signal processor (DSP) 〇I 0. Such as the scope of patent application No. 9 In the device for realizing the caller's incoming call and voice, the digital signal processor is a DSP unit in a baseband chip of the mobile communication device. II. A method for realizing the caller's incoming message and voice implementation, which can be applied to a mobile communication device, and is used to prompt the above-mentioned mobile communication device for an incoming call message by voice. It includes the following steps: Provide a contact database to record The caller's name and its corresponding caller's phone number 'also provide a voice database to record the phoneme unit and its corresponding phoneme data; when the mobile communication device receives an incoming call, retrieve the caller's phone number of the incoming call; According to the caller's phone number of the above incoming call, the corresponding caller name is queried through the address book database; according to the corresponding caller name obtained through the query, the corresponding caller name is queried through the voice database-the plural number included A phoneme data; using a voice processing unit to synthesize the first phoneme data obtained by the query to generate a first sound signal; and playing the first sound signal. ·-12 * Caller incoming call message as described in item 11 of the scope of patent application Page 19, 42417 ^^ 'I-VI. Processing method of patent application tone realization' which further includes the following steps: receiving one of the short messages received by the above mobile communication device from the outside; querying the above through the above voice database Plural second phoneme data corresponding to the short message; 11 using the above-mentioned voice processing unit to synthesize the above-mentioned second phoneme data to generate a second sound signal; and playing the above-mentioned second sound signal D 1 3 The method for realizing the caller's incoming call voice described in the above item further includes the following steps: When the mobile communication device receives the incoming call, take out one of the preset incoming call texts; query the incoming call texts through the above voice database Corresponding plural third phoneme data; using the voice processing unit, synthesizing and generating the third phoneme data—a third sound signal; and playing the third sound signal. 14. The method for processing the caller's incoming sfi tone as described in item 11 or 12 or 13 of the scope of the patent application, further comprising a step of displaying the above-mentioned caller name inquired on a liquid crystal display screen. 15. The method for realizing the caller's incoming message voice according to item 11 or 12 or 13 of the patent application park, wherein the above-mentioned voice database is a database established by using Text To Speech technology. 16 · The method for realizing electric message voice according to the calling party described in item n * 12 or 13 of the scope of patent application, wherein the above-mentioned voice database is for profit Page 20 Ο 3 Page 21