JP6249760B2 - Text-to-speech device - Google Patents

Description

  The present invention relates to a text reading device that reads text.

  2. Description of the Related Art Conventionally, a text-to-speech apparatus having a TTS (Text To Speech) function has been developed that converts text such as sentences or words contained in content (for example, dictionary content and book content) into speech and reads it out. An example of such a text-to-speech device is an electronic dictionary equipped with a foreign language TTS engine that converts foreign language text such as English into speech. Furthermore, there is a text-to-speech device equipped with a Japanese TTS engine that converts Japanese text into speech for Japanese learners and those who wish to read by the device.

  Patent Document 1 discloses an example of a text-to-speech device equipped with a Japanese TTS engine. This text-to-speech device obtains the kana character string reading by referring to the dictionary data part for kana-kanji conversion when the kanji character string to be read is not registered in the speech conversion dictionary data part. . Then, the sound corresponding to the kana is reproduced instead of the sound corresponding to the kanji character string.

JP 2000-194389 A (published July 14, 2000)

  In a conventional text-to-speech device equipped with a Japanese TTS engine, including the text-to-speech device disclosed in Patent Document 1, when there are a plurality of readings in a kanji character string in a text, the Japanese TTS engine determines one The kanji character string is read out only by reading. Therefore, when the kanji character string has a plurality of readings, there is a possibility that the kanji character string is read out in a reading that does not conform to the intention of the content creator. That is, the conventional text-to-speech device has a problem that the text-to-speech accuracy including the kanji character string is low.

  In addition, the text may include a character string in a language such as English or Chinese as well as Japanese. In this case, it is desirable that each language character string be read out smoothly.

  However, in general, one TTS engine is not adapted to read multiple language strings. As an exception, the Japanese TTS engine can read not only Japanese character strings but also English character strings. But you can't read Chinese strings. For this reason, the conventional text-to-speech device has a problem that it cannot smoothly read a text in which a plurality of languages are mixed.

  The present invention has been made in view of the above problems, and an object of the present invention is to provide a text-to-speech device that can smoothly read a text in which a plurality of languages are mixed.

  In order to solve the above-described problem, a text-to-speech device according to one aspect of the present invention converts text data to be displayed on a display unit into voice data, and outputs the converted voice data. A plurality of speech data generation units that respectively generate speech data in one or more languages, determine a language type of a character string within a range of the selected text data, and A selection unit that selects one voice data generation unit capable of generating voice data of character strings of a plurality of types of languages included in the text data from a plurality of voice data generation units.

  According to one aspect of the present invention, it is possible to smoothly read a text in which a plurality of languages are mixed.

It is a block diagram which shows the structure of the control part with which the information processing apparatus which concerns on one Embodiment of this invention is provided. It is a block diagram which shows the structure of the information processing apparatus which concerns on one Embodiment of this invention. (A) is a figure which shows an example of the content data which concerns on one Embodiment of this invention, (b) is extracted from the content data which concerns on one Embodiment of this invention, and is stored in the reading DB part It is a figure which shows the data format of a character string and its reading kana. It is a flowchart which shows the operation | movement at the time of reading-out data processing of the information processing apparatus which concerns on one Embodiment of this invention. It is an example of a display of the content in the display part with which the information processing apparatus which concerns on one Embodiment of this invention was provided. It is a part of flowchart which shows the flow of the reading-out data process which the control part of the information processing apparatus which concerns on one Embodiment of this invention performs. FIG. 6 is a part of a flowchart showing a flow of read-out data processing executed by a control unit according to an embodiment of the present invention, and is the remainder of the flowchart shown in FIG. (A) And (b) is an enlarged view of the information displayed on the display part of the information processing apparatus which concerns on one Embodiment of this invention. (A) And (b) is the example of a display of the display part in one modification of this invention. (A)-(c) is a figure which shows the example of a display of the text with ruby information which concerns on one Embodiment of this invention, (a) And (c) is a table | surface with which the ruby of a text is non-displaying. It is a display example, (b) is a display example in which ruby is displayed on the kanji character string in the text. It is a block diagram which shows the structure of the control part with which the information processing apparatus which concerns on other embodiment of this invention is provided. It is an example of a display of the content in the display part with which the information processing apparatus which concerns on other embodiment of this invention was provided. It is another example of the display of the content in the display part with which the information processing apparatus which concerns on other embodiment of this invention was provided. It is the further another example of a display of the content in the display part with which the information processing apparatus which concerns on other embodiment of this invention was provided. It is a flowchart which shows the flow of the reading-out data process which the control part of the information processing apparatus which concerns on other embodiment of this invention performs.

Embodiment 1
Hereinafter, an embodiment of the present invention will be described in detail.

[Configuration of Information Processing Apparatus 1]
The information processing apparatus 1 (text reading apparatus) according to the present embodiment is an apparatus that reads a text included in content. First, the configuration of the information processing apparatus 1 according to the present embodiment will be described with reference to FIG. FIG. 2 is a block diagram illustrating a configuration of the information processing apparatus 1.

  As illustrated in FIG. 2, the information processing apparatus 1 includes a display unit 10, a touch panel 20, a TTS engine 30 (kanji character string voice data generation unit), a control unit 40, a storage unit 50, a temporary storage unit 60, and a speaker 70. I have.

  The control unit 40 integrally controls each unit (including the display unit 10, the touch panel 20, the TTS engine 30, and the speaker 70) included in the information processing apparatus 1.

  The storage unit 50 includes a content storage unit 50A and a reading DB unit 50B. Content data such as an electronic dictionary or an electronic book is stored in the content storage unit 50A. The content data includes image data or text data. In the reading DB 50B, the control unit 40 stores data in which the kanji character strings included in the text data and the readings are associated with each other in a database. The text data may have tag information that associates a kanji character string in the text data and a kana character string corresponding thereto with each other as a kanji character string and its reading. A specific example of tag information will be described later.

  Text data is temporarily stored in the temporary storage unit 60 by the control unit 40. More specifically, the control unit 40 acquires text data from the content stored in the content storage unit 50A and stores the text data in the temporary storage unit 60. Or the control part 40 acquires the text data of the range selected by the user among the whole text data displayed on the display part 10, and memorize | stores it in the temporary memory part 60. FIG. The temporary storage unit 60 may be included in the storage unit 50.

  On the display unit 10, data presented to the user by the control unit 40, for example, image data or text data included in the content is displayed. Specifically, the content data acquired by the control unit 40 from the content storage unit 50A is displayed. The touch panel 20 is provided on the display unit 10, detects an operation performed by the user touching the display unit 10, and outputs a detection signal as a result thereof to the control unit 40. The control unit 40 detects a touch on the display position of the text data based on a detection signal from the touch panel 20.

  Data obtained by converting the text data stored in the temporary storage unit 60 by the control unit 40 is input to the TTS engine 30. The TTS engine 30 converts the text data input from the control unit 40 into voice data. Then, the TTS engine 30 outputs the converted audio data to the control unit 40. From the speaker 70, sound based on the sound data acquired by the control unit 40 from the TTS engine 30 is output.

  FIG. 3A shows an example of text data having tag information. FIG. 3B shows the data format of information stored in the reading DB unit 50B. In the text data shown in FIG. 3A, tag information (<kanji>, </ kanji>, <yomi>, and </ yomi>) that is not displayed when content data is reproduced is referred to as “miscellaneous song”. A kanji character string and a kana character string “Elephant” are associated as a kanji character string and its reading. When the control unit 40 processes the text data shown in FIG. 3A, based on the tag information, the control unit 40 starts and ends the kanji character string “miscellaneous” included in the text data, and “Zooka”. Is detected from the text data. Then, as shown in FIG. 3B, the kana character string “Elephant” is associated with the start position and end position of the kanji character string “miscellaneous” as a reading and stored in the reading DB unit 50B. Let

[Detailed Configuration of Control Unit 40]
Next, a detailed configuration of the control unit 40 will be described with reference to FIG. FIG. 1 is a block diagram illustrating a configuration of the control unit 40. As shown in FIG. 1, the control unit 40 includes a kanji extraction unit 42 (kanji character string extraction unit), a reading kana extraction unit 43 (kana character string extraction unit), and a kanji kana character replacement unit 44 (kanji character string speech data generation unit). ), A TTS engine output unit 45, and a speaker output unit 46.

  The kanji extraction unit 42 extracts a kanji character string included in the text data. The reading kana extraction unit 43 extracts the kanji character string extracted by the kanji extraction unit 42 and the reading kana associated with the text data from the text data. The reading kana extraction unit 43 includes a reading DB reference extraction unit 430, a parenthesis character string extraction unit 431, and a reading kana determination unit 432. The reading DB reference extraction unit 430 refers to the reading DB unit 50B and extracts the reading kana of the kanji character string in the text data from the reading DB unit 50B. Specifically, the reading DB reference extraction unit 430 acquires the start position and the end position indicating the position of the kanji character string in the text data from the reading DB shown in FIG. To extract the readings associated with the kanji strings to be played.

  The parenthesis character string extraction unit 431 includes a kana character string (a hiragana character string or a katakana character string) enclosed in parentheses (open and close parentheses) immediately after the kanji character string extracted by the kanji extraction unit 42. If so, the kana character string is extracted. The parenthesis character string extraction unit 431 determines whether there is a kana character string enclosed in parentheses and parentheses immediately after the kanji character string for the kanji character string from which the reading kana character extraction is performed by the reading DB reference extraction unit 430. The determination of whether or not may be omitted. The reading kana determination unit 432 determines whether the parenthesis character string extracted by the parenthesis character string extraction unit 431 is a reading kana of the kanji character string immediately before the parenthesis.

  For example, when a character string “Zoka” (see FIG. 5) exists in the text data, the character string extraction unit 431 in parentheses stores the kanji character string “Zaika” included in this character string. The kana character string “zouka” surrounded by parentheses “()” immediately after is extracted as a reading kana for the kanji character string “miscellaneous”. In other words, the parenthesis character string extraction unit 431 extracts the kana character string (“Elephant”) corresponding to the kanji character string (“miscellaneous song”). The reading kana extraction unit 43 replaces the kana character string enclosed in parentheses “()” with “<>”, “<< >>”, or other specific symbols. May be extracted. In other words, the reading kana extraction unit 43 may extract a kana character string surrounded by a predetermined first symbol and a predetermined second symbol corresponding to the predetermined first symbol. Here, the first symbol and the second symbol may be the same symbol, or may be a pair of symbols such as parenthesis opening and parenthesis closing, or irrelevant to each other. (For example, $ and #). However, which symbol corresponds to the first symbol or the second symbol is defined in advance in the information processing apparatus 1.

  The kanji kana replacement unit 44 performs a replacement process on the text data stored in the temporary storage unit 60 based on the result of the reading kana extraction by the reading kana extraction unit 43. Specifically, the kanji kana replacement unit 44 extracts the kanji character string from which the reading kana is extracted by the reading DB reference extraction unit 430 from the kanji character strings in the text data stored in the temporary storage unit 60. Replace with the reading. Further, the kanji kana replacement unit 44 uses the kana character string extracted by the kanji extraction unit 42 from the replaced text data as the reading kana of the kanji character string by the reading kana determination unit 432. Replace with. The TTS engine output unit 45 outputs the text data after the kanji kana replacement processing by the kanji kana replacement unit 44 to the TTS engine 30.

  The TTS engine 30 converts the text data input from the TTS engine output unit 45 into voice data. Thereafter, the TTS engine 30 outputs the converted audio data to the speaker output unit 46. The speaker output unit 46 converts the audio data input from the TTS engine 30 into an electrical signal and outputs it to the speaker 70.

[Operation of Control Unit 40]
Next, the operation of the control unit 40, particularly, the operation of the control unit 40 related to reading data processing will be described. In the reading data processing, the information processing apparatus 1 reads the text data. Below, the outline | summary of operation | movement of the control part 40 until read-out data processing is performed first is demonstrated. Thereafter, details of the reading data processing will be described.

[Operation of control unit 40; operation of control unit 40 until read-out data processing is executed]
The outline of the operation of the control unit 40 until the reading data processing is executed will be described with reference to FIGS. 4 and 5. FIG. 4 is a flowchart showing the operation of the control unit 40 until the reading data processing is executed. FIG. 5 is a display example of content on the display unit 10. In the following description, actually touching the touch panel 20 is expressed as touching the display unit 10.

  As shown in FIG. 4, the control unit 40 first determines whether or not content is being displayed on the display unit 10 (S1). When the content is being displayed (Yes in S1), the control unit 40 determines whether or not all or part of the text displayed on the display unit 10 has been touched by the user (S2). Hereinafter, all or part of the touched text is referred to as selected text. In FIG. 5, “Miscellaneous Song” which is the selected text is displayed with a shadow on the display unit 10. The selected text may be selected using the touch pen P as shown in FIG. 5 or may be performed with the user's finger. In addition, the selected text may be selected by the user operating a key (not shown) provided in the information processing apparatus 1.

  When the content is not being displayed (No in S1) or all or part of the text is not touched (No in S2), the control unit 40 ends the operation without performing the reading data processing. On the other hand, when the content is being displayed (Yes in S1) and all or part of the text included in the content is touched (Yes in S2), as shown in FIG. The touch menu TM is displayed (S3). As shown in the figure, the touch menu TM includes an option of “speech” that causes the information processing apparatus 1 to read the selected text.

  The touch menu TM may include, in addition to “speech”, “markers” displayed by drawing a marker on the selected text (see FIG. 5). In addition, the touch menu TM may include, for example, an option for performing a dictionary search for terms included in the selected text.

  Note that the operation that the user performs to select text is not limited to the touch operation. For example, when a button is displayed on the screen of the display unit 10 and an operation of pressing this button is performed, an operation in accordance with specifications and settings previously determined in the information processing apparatus 1 (for example, during display) The selected text may be determined by this operation.

  After displaying the touch menu TM (after S3), the control unit 40 determines whether or not “read” is selected from the touch menu TM (S4). When an option other than “Read” is selected (No in S4), the control unit 40 performs an operation corresponding to the selected option, and then ends the operation without performing the read data processing.

  On the other hand, when “Reading” is selected (Yes in S4), the control unit 40 executes reading data processing (S5). In the read-out data process S5, the control unit 40 executes the kanji kana replacement process for the selected text, and then outputs the selected text subjected to the kanji kana replacement process to the TTS engine 30. Details of the reading data processing S5 will be described later. In the TTS engine 30, the text data after the kanji-kana replacement process is converted into voice data. Thereafter, the control unit 40 acquires the audio data output from the TTS engine 30 and outputs it to the speaker 70. Thereafter, the process ends.

  Although the configuration for displaying the touch menu TM has been described here, the touch menu TM is displayed when there is only one predetermined operation (that is, execution of the reading process) triggered by the text selection operation. The reading process may be executed without any problem.

[Operation of control unit 40; details of reading data processing S5]
Next, details of the reading data processing S5 described above will be described with reference to FIGS. 6 and 7 are flowcharts showing the flow of the read-out data processing S5.

  In the reading data processing S5, first, the control unit 40 copies the selected text to the temporary storage unit 60 (S501). In subsequent steps, each unit of the control unit 40 performs a process of replacing the kanji character string with the kana character string for the selected text in the temporary storage unit 60.

  After step S501, the reading DB reference extraction unit 430 refers to the reading DB unit 50B to determine whether or not the selected text includes a kanji character string associated with the reading kana (S502). When the selected text includes a kanji character string associated with the reading kana (Yes in S502), the kanji kana replacement unit 44 replaces the kanji character string with the reading kana (S503).

  When the kanji character string associated with the reading kana is not included in the selected text (No in S502), or after step S503, the kanji extraction unit 42 determines whether the selected text includes a kanji character string. Is determined (S504). If the selected text does not contain a Kanji character string (No in S504), the TTS engine output unit 45 outputs the selected text in the temporary storage unit 60 to the TTS engine 30 (S511), and the read-out data processing S5 ends. To do.

  On the other hand, when the selected text includes a kanji character string (Yes in S504), the character string extraction unit 431 in parentheses immediately follows the kanji character string in the selected text (hereinafter referred to as a character string enclosed in parentheses). Then, it is determined whether or not there is a character string in parentheses (S505).

  If there is no character string in parentheses immediately after the kanji character string in the selected text (No in S505), the character string extraction unit 431 in parentheses determines whether the end of the selected text is a kanji character string. (S506). If the end of the selected text is not a Chinese character string (No in S506), the TTS engine output unit 45 outputs the selected text in the temporary storage unit 60 to the TTS engine 30 (S511), and the read-out data processing S5 ends. .

  On the other hand, when the end of the selected text is a Kanji character string (Yes in S506), the parenthesized character string extraction unit 431 determines whether or not the parenthesized character string exists immediately after the selected text (S507). . When there is no character string in parentheses immediately after the selected text (No in S507), the TTS engine output unit 45 outputs the selected text in the temporary storage unit 60 to the TTS engine 30 (S511), and the read-out data processing S5 finish.

  In the selected text, if there is a character string in parentheses immediately after the kanji character string (in the example shown as Yes or Ex1 in S505) or immediately after the selected text that ends with the kanji character, Is present (in the case of the example shown as Yes and Ex2 in S507), the reading kana determination unit 432 determines whether or not the character string in parentheses is a kana character string consisting only of hiragana or only katakana (S508). . Thereby, it is determined whether or not the character string in parentheses is a representation of the reading of the immediately preceding kanji character string.

  When the character string in the parenthesis includes a character string other than hiragana and katakana (in the example shown as No and Ex3 in S508), the TTS engine output unit 45 outputs the selected text in the temporary storage unit 60 to the TTS engine 30. (S511), the reading data processing S5 ends.

  On the other hand, if the character string in parentheses is a kana character string made up of hiragana or katakana (Yes in S508), the reading kana determination unit 432 determines whether the length of the character string in parentheses is within a predetermined range. (S509, kana character string extraction step). The predetermined range may be within three times the character string length of the kanji character string immediately before the parentheses. Thus, the reading character determination unit 432 determines that the character string in parentheses is not the reading character of the kanji character string when the character string in parentheses is too long as the reading character name of the kanji character string immediately before the parenthesis. can do.

  When the length of the character string in the parenthesis is not within the predetermined range (No in S509), the TTS engine output unit 45 outputs the selected text in the temporary storage unit 60 to the TTS engine 30 (S511), and the read-out data processing S5 finish. On the other hand, when the length of the character string in the parenthesis is within the predetermined range (Yes in S509), the kanji kana replacement unit 44 replaces the character string in front of the parenthesis with the character string in parenthesis, and further opens the parenthesis. If part or all of the character string up to closing the parenthesis is included in the selected text, the character string is deleted from the selected text (S510). Thereafter, the TTS engine output unit 45 outputs the selected text in the temporary storage unit 60 to the TTS engine 30 (S511).

  Here, in step S510, when the kanji character string immediately before the parenthesis is composed of one kanji, the kanji kana replacement unit 44 replaces the one kanji with the character string in parentheses. On the other hand, when the kanji character string immediately before the parenthesis is composed of a plurality of kanji characters, the kanji kana replacement unit 44 replaces the plurality of kanji characters with the character string in parentheses. In other words, the kanji kana replacement unit 44 traces the text immediately before opening the parenthesis and replaces a series of kanji characters composed of kanji having a kanji character code with a character string in parentheses.

  Thus, even if the selected text is an example shown in Ex1 including parentheses (“Miscellaneous songs”) or an example where the selected text is shown in Ex2 not including parentheses (“Miscellaneous songs”). However, the selected text “Elephant” shown in Ex4 is input to the TTS engine 30. Thus, the reading data processing S5 is completed.

  Although not shown, after step S511, the TTS engine 30 to which the selected text data is input converts the selected text data into voice data and outputs the voice data to the control unit 40 (voice data generation step). The control unit 40 outputs the audio data input from the TTS engine 30 to the speaker 70. The speaker 70 outputs sound based on the sound data input from the control unit 40. As a result, the kanji character string is read out in the reading as intended by the creator of the content.

  Here, an example has been described in which the information processing apparatus 1 reads the selected text selected by the user, but the present invention is not limited to this. By adopting a configuration in which the information processing device 1 sequentially reads the entire text data instead of the selected text, a configuration in which the information processing device 1 reads the text can be realized.

[Modification of Reading Data Processing S5]
The control unit 40 may perform only kanji kana replacement using a character string in parentheses without performing kanji kana replacement with reference to the reading DB 50B. In this case, step S502 and step S503 are not executed in the reading data processing S5. That is, step S504 is executed following step S501.

  Further, in the reading data processing S5, step S503 may be executed at other timing as long as it is executed before step S511 in which the selected text is output to the TTS engine 30. That is, step S503 may be executed between step S504 and step S510, or may be executed subsequent to step S510.

  Furthermore, in the read-out data processing S5, if Yes in step S509, that is, if the length of the parenthesized character string is within a predetermined range, the kanji kana replacement unit 44 outputs the parenthesized character string to the TTS engine output unit 45. May be. In the case of this configuration, the TTS engine output unit 45 outputs a parenthesized character string to the TTS engine 30 instead of the selected text. The TTS engine 30 converts the character string in parentheses into voice data and outputs it to the control unit 40. The control unit 40 outputs the audio data input from the TTS engine 30 to the speaker 70. As a result, the parenthesis character string is output from the speaker 70 instead of the kanji character string included in the selected text. This also has the effect of reading out a kanji character string as intended by the creator of the content.

[Comparison with conventional text-to-speech devices]
Here, a specific effect achieved by the information processing apparatus 1 according to the present embodiment will be described. Therefore, referring to FIGS. 8A and 8B, the operation of the information processing apparatus 1 according to the present embodiment and the operation of a conventional text-to-speech apparatus will be compared. FIGS. 8A and 8B are both enlarged views of the display unit 10 of the information processing apparatus 1. In (a) of FIG. 8, the kanji character string "Miscellaneous song" is the selected text. On the other hand, in FIG. 8B, the kanji character string (“Miscellaneous Song”), followed by the parenthesis and the hiragana character string (“(Elephant)”) within the parenthesis are the selected text.

  As described above, in the conventional text-to-speech device, the TTS engine selects the reading of the kanji character string in the text. Therefore, when a Kanji character string has a plurality of readings (for example, “Okami” can be read as “Okami”, “Dajin”, or “Oomiwa”), the TTS engine selects any one of the readings. . Therefore, there is a possibility that a reading different from the intention of the content creator is selected. For example, even though the content creator created the content with the intention of reading “Okami” as “Oomiwa”, the TTS engine would select “Okami” as the reading of “Okami”. is there.

  In the conventional text-to-speech device, as shown in FIG. 8B, when the selected text includes a kanji character string and a reading kana of the kanji character string, the kanji is different from the intention of the content creator. There is a possibility that both the character string and its kana will be read out. For example, the content creator may have read the selected text “Zoka” simply as “Zoka” but it reads “Zozuka”. That's right.

  On the other hand, in the information processing apparatus 1 according to the present embodiment, as shown in FIGS. 8A and 8B, immediately after the kanji character string (“Miscellaneous Song”), it is enclosed between the parenthesis opening and the parenthesis closing. If the read kana (“Zouka”) exists, the result of the reading data processing S5 is that the character string from the Kanji character string to the parenthesis closing (“Zoka”) is replaced with the above A reading pseudonym (“Elephant”) is input to the TTS engine 30. Therefore, the information processing apparatus 1 can read the kanji character string by reading as intended by the content creator, unlike the conventional text reading apparatus. Further, when the selected text is “Miscellaneous Song”, the information processing apparatus 1 obtains the character string “Elephant” from the selected text in the above-described reading data processing S5, and uses this character string as the TTS engine. Output to 30. Therefore, the selected text is read as “Elephant” as intended by the creator of the content.

[Modification 1]
FIG. 9A shows a display example of the display unit 10 in a modification of the present embodiment. In the first modification, as shown in FIG. 9A, the touched character string (“Miscellaneous Song”) is a Kanji character string, and parentheses (opening and closing parentheses) are immediately after the Kanji character string. If it exists, the control unit 40 uses the touched kanji character string to the character string ("Zozuka") from the touched kanji character string to the closing of the parenthesis immediately after the kanji character string.

  In the configuration of the first modification, the control unit 40 passes the kana character (“Elephant”) in parentheses to the TTS engine 30 when “read” by the TTS engine 30 is selected from the touch menu TM or the like. On the other hand, when “dictionary search” is selected, the control unit 40 does not adopt the character string (“(Elephant)”) from the opening parenthesis to the closing parenthesis as the search keyword. That is, when the dictionary search is selected, the control unit 40 adopts only the touched character string (“miscellaneous song”) among the selected text as a search keyword. As a result, when a dictionary search is selected after a character string is touched, the character string from the parenthesis opening to the parenthesis closing not included in the touched character string becomes the dictionary search range by the dictionary search engine. It can prevent limiting to unnecessary.

[Modification 2]
FIG. 9B shows a display example of the display unit 10 in another modification of the present embodiment. In the second modification, the control unit 40 displays a pop-up menu PUM near the touched character string (“Miscellaneous Song”) as shown in FIG. 9B. Then, the user selects a character string to be selected text from the touched character string from the pop-up menu PUM.

  In the configuration of the second modification, the user can determine a character string to be selected text. As a result, it is possible to realize reading of a character string in an arbitrary range included in the touched character string, as desired by the user.

[Modification 3]
In the present embodiment, the configuration in which text data read by the information processing apparatus 1 is displayed on the display unit 10 has been described. However, in the modification, the text data read out by the information processing apparatus 1 may not be displayed on the display unit 10. For example, when the user selects a file in the content storage unit 50A, the information processing apparatus 1 may read the text data in the file without displaying it.

[Embodiment 2]
Another embodiment of the present invention will be described below with reference to FIGS. 10A to 10C. For convenience of explanation, members having the same functions as those described in the embodiment are given the same reference numerals, and descriptions thereof are omitted.

  The text data included in the content may be preliminarily provided with ruby information that associates a kanji character string in the text data with a ruby (a hiragana character string or a katakana character string) representing the reading of the kanji character string. is there. FIGS. 10A to 10C show display examples of text with ruby information. FIGS. 10A and 10C are display examples of text with ruby information in which ruby Ru is not displayed. FIG. 10B is a display example of text with ruby information in which ruby Ru is displayed on the kanji character string.

  The control unit 80 provided in the information processing apparatus 2 of the present embodiment includes a reading kana extraction unit 83 (a kana character string extraction unit) instead of the reading kana extraction unit 43 of the control unit 40 of the embodiment. The storage unit 50 further stores ruby information. Other configurations of the information processing apparatus 2 are the same as those of the information processing apparatus 1.

  When the hiragana character string or the katakana character string enclosed in parentheses exists immediately after the kanji character string in the text data, the reading kana extraction unit 43 according to the first embodiment extracts the hiragana character string or the katakana character string. , Extracted as a kana character string associated with the kanji character string. On the other hand, the reading kana extraction unit 83 according to the present embodiment refers to the ruby information and extracts the kanji character string ruby Ru as the kana character string associated with the kanji character string.

[Operation of control unit 80]
Below, with reference to FIG.10 (c), operation | movement of the control part 80 which concerns on this embodiment is demonstrated. As shown in (c) of FIG. 10, the control unit 80 determines a character string touched by the touch pen P or the like (“lower person”) as the selected text. And the control part 40 displays the touch menu TM for selecting the operation | movement with respect to the touched character string.

  When an operation of selecting a “speech” option from the touch menu TM is performed, the reading kana extraction unit 83 refers to the ruby information and is associated with the kanji character string (“sludge”) in the selected text. Extract Ruby Ru ("Genin"). The kanji-kana replacement unit 44 replaces the kanji character string (“lower person”) in the selected text with the ruby Ru (“genin”) extracted by the reading-kana extraction unit 83. The TTS engine output unit 45 outputs the selected text (“Genin”) after the replacement process by the kanji kana replacement unit 44 to the TTS engine 30. After the selected text input to the TTS engine 30 is converted into voice data, voice output is performed from the speaker 70 based on the voice data.

  The reading kana extraction unit 83 extracts the kana character string associated with the kanji character string according to whether or not the text is text with ruby information and the method of the present embodiment. You may change between Specifically, when the ruby information is added to the text, the reading kana extraction unit 83 associates the ruby of the kanji character string with the kanji character string based on the ruby information as in the present embodiment. The extracted kana character string is extracted. On the other hand, when the ruby information is not given to the text, the hiragana or katakana character string in parentheses immediately after the kanji character string is associated with the kanji character string as in the first embodiment. Extract as a kana character string.

[Embodiment 3]
The following will describe another embodiment of the present invention with reference to FIGS. For convenience of explanation, members having the same functions as those described in the above embodiments are denoted by the same reference numerals and description thereof is omitted.

  A TTS engine (hereinafter referred to as a Japanese TTS engine) for reading out Japanese, such as the TTS engine 30 (see FIG. 1) of the first embodiment, is generally used in addition to Japanese speech data. The voice data for each letter of the alphabet can be output. Therefore, the Japanese TTS engine can read out not only Japanese character strings but also English character strings. When the Japanese TTS engine reads out English, the Japanese TTS engine outputs a combination of voice data of each alphabet constituting the English. Therefore, the Japanese TTS engine cannot read English smoothly with a natural pronunciation like native.

  However, the Japanese TTS engine cannot read out characters in languages other than Japanese and English (for example, Chinese).

  Therefore, in the present embodiment, an information processing apparatus including a plurality of TTS engines including a Japanese TTS engine and configured to smoothly read out languages other than Japanese will be described. The information processing apparatus according to the present embodiment selects any TTS engine from a plurality of TTS engines based on the language of the character string in the text. The text is read out using the selected TTS engine. The information processing apparatus according to the present embodiment executes the reading data processing (see FIGS. 6 and 7) described in the first embodiment, particularly when the Japanese text is read out using the Japanese TTS engine.

  Note that the information processing apparatus of the present embodiment may be the information processing apparatus 1 of the first embodiment or the information processing apparatus 2 of the second embodiment.

[Configuration of the information processing apparatus of this embodiment]
The configuration of the information processing apparatus according to this embodiment will be described with reference to FIG. FIG. 11 is a configuration diagram illustrating a configuration of the control unit 940 provided in the information processing apparatus of the present embodiment. The information processing apparatus according to the present embodiment includes a control unit 940 instead of the control unit 40 in the configuration of the information processing apparatus 1 according to the first embodiment. Further, in place of the TTS engine 30 of the information processing apparatus 1, a Japanese TTS engine 931 (voice data generation unit, first voice data generation unit), an English TTS engine 932 (voice data generation unit), and a Chinese TTS engine 933. (Audio data generation unit, second audio data generation unit). The Japanese TTS engine 931 can read Japanese and English (alphabet). The English TTS engine 932 can read only English. In addition, the Chinese TTS engine 933 can read out Chinese and English (alphabet). The Japanese TTS engine 931 has the same function as the TTS engine 30 of the first and second embodiments, and can execute the same process as the process of the TTS engine 30 described in the first and second embodiments. it can.

[Detailed Configuration of Control Unit 940]
As shown in FIG. 11, the control unit 940 further includes a language determination unit 941 (selection unit) in addition to all the components of the control unit 40 of the first embodiment.

  The language determination unit 941 acquires the text selected by the user, that is, the selected text, from the temporary storage unit 60. Then, the language type of the character string in the selected text is determined. In addition, the language determination unit 941 reads out the selected text from the Japanese TTS engine 931, the English TTS engine 932, and the Chinese TTS engine 933 when character strings of a plurality of types of languages exist in the selected text. A selection engine determination process for selecting an engine (hereinafter referred to as a selection engine) is executed.

  Specifically, in the selection engine determination process, the language determination unit 941 first determines how many words the first character of the selected text is. That is, the language determination unit 941 determines the language type of the first character. And the information (language setting) which shows the language of the determined character is hold | maintained. Next, it is determined how many words the second character of the selected text is. Then, the language setting is updated from information indicating the language of the first character to information indicating the language of the second character. Hereinafter, in the same manner, the characters of the selected text are sequentially determined one by one in order of the number of the characters, and the language setting is updated based on the determination result.

  The language determination unit 941 determines how many words the last character of the selected text is, and when the language setting is updated based on the determination result, the selection engine is determined according to the language setting at that time. . For example, when the language setting at the above time is Japanese, the language determination unit 941 selects the Japanese TTS engine 931 as the selection engine. In addition, when the language setting at the time is English or Chinese, the language determination unit 941 selects the English TTS engine 932 and the Chinese TTS engine 933 as selection engines.

  According to the selection engine determination process, for example, in the selected text, the first part is an English character string, the middle part is a Japanese character string, and the last part is a Chinese character string. If so, the language determination unit 941 selects the Japanese TTS engine 931.

  As described above, the language determination unit 941 includes, from the Japanese TTS engine 931, the English TTS engine 932, and the Chinese TTS engine 933, in the selected text when character strings of a plurality of types of languages exist in the selected text. One TTS engine capable of reading out character strings in a plurality of types of languages is selected.

  By the way, as described above, the Japanese TTS engine 931, the English TTS engine 932, and the Chinese TTS engine 933 each have characters that cannot be read out. Such characters are referred to herein as external characters. For example, for the Japanese TTS engine 931, characters in languages other than Japanese and English (eg, Chinese) correspond to external characters. For the Chinese TTS engine 933, characters in languages other than Chinese and English (for example, Japanese) correspond to external characters.

  Therefore, when the selected text includes an external character, the language determination unit 941 sets the character string from the first character of the selected text to the character immediately before the external character as a new selected text. In other words, the language determination unit 941 updates the selected text so that the selection engine can read out the entire updated selected text. In addition, the language determination unit 941 may use a character string from the first character of the selected text to the character immediately before the line feed as a new selected text.

  When the Japanese TTS engine 931 is selected as the selection engine, the language determination unit 941 outputs the updated selected text to the kanji extraction unit 42. When the English TTS engine 932 or the Chinese TTS engine 933 is selected as the selection engine, the updated selected text and language setting are output to the TTS engine output unit 45.

  The TTS engine output unit 45 outputs the selected text input from the language determination unit 941 to the English TTS engine 932 or the Chinese TTS engine 933 according to the language setting input from the language determination unit 941. Specifically, the TTS engine output unit 45 outputs the selected text to the English TTS engine 932 when the language setting is English. If the language setting is Chinese, the selected text is output to the Chinese TTS engine 933. Since the process executed by the kanji extraction unit 42 to which the selected text is input has been described in the first embodiment, description of the process is omitted.

[How to select a selection engine]
Here, a method in which the language determination unit 941 determines a selection engine from a plurality of TTS engines will be specifically described.

  The language determination unit 941 selects a TTS engine that can read out the longest selected text range from the Japanese TTS engine 931, the English TTS engine 932, and the Chinese TTS engine 933 as the selection engine.

  For example, when the selected text includes Japanese and English, the language determination unit 941 selects the Japanese TTS engine 931 instead of the English TTS engine 932 or the Chinese TTS engine 933 as the selection engine. This is because the English TTS engine 932 and the Chinese TTS engine 933 can read only the English in the selected text. On the other hand, the Japanese TTS engine 931 can read both Japanese and English in the selected text. Therefore, the Japanese TTS engine 931 can read out a longer selected text range than the English TTS engine 932 and the Chinese TTS engine 933.

  Similarly, when the selected text includes Chinese and English, the language determination unit 941 selects the Chinese TTS engine 933 instead of the Japanese TTS engine 931 or the English TTS engine 932 as the selection engine.

  When the selected text includes Japanese and Chinese, the language determination unit 941 selects, as the selection engine, a TTS engine that can read a language that appears before the other language in the selected text. For example, when Japanese and Chinese appear in this order in the selected text, the language determination unit 941 selects the Japanese TTS engine 931 as the selection engine.

  When the selected text includes English but does not include Japanese and Chinese, the language determination unit 941 selects the English TTS engine 932 as the selection engine. This is because the English TTS engine 932 can read English naturally and more smoothly than the Japanese TTS engine 931 and the Chinese TTS engine 933.

  The method for determining the selection engine is not limited to the method described above. For example, in another configuration, the language determination unit 941 uses a Japanese TTS engine 931, an English TTS engine 932, and a Chinese as selection engines that read out a Japanese part, an English part, and a Chinese part in the selected text, respectively. A TTS engine 933 may be selected.

  In this configuration, Japanese, English, and Chinese in the selected text are all read out. Furthermore, in the selected text, each part of Japanese, English, and Chinese is read out by a TTS engine adapted to read out the language, so that each of the above parts is read out naturally and smoothly.

[Specific example of reading aloud]
A specific example in which the selected text is read out by the information processing apparatus according to the present embodiment will be described with reference to FIGS. 12 to 14 are diagrams each showing the display unit 10 (see FIG. 2) provided in the information processing apparatus of the present embodiment. In each of the above drawings, a part of the text displayed on the screen of the display unit 10 is selected with the touch pen P, and a touch menu showing processing that can be executed on the selected text (selected text). TM is displayed.

  Selected text shown in Figure 12

Consists of Japanese, English, Chinese and symbols. The language determination unit 941 selects the Japanese TTS engine 931 as a selection engine that reads out the selected text. The beginning of the character string read out by the selection engine is “Today” and the end is “Hi!”. In the selected text shown in FIG. 12, the Chinese after “Hi!” Is not read out because it is an external character for the Japanese TTS engine 931.

  Here, the Japanese TTS engine 931 executes the reading data processing (see FIGS. 6 and 7) described in the first embodiment when reading the Japanese character string “Today”. Then, instead of the kanji character string “today”, the kana character string “Konichi” is read out.

  The selected text “Hello! / Hi!” Shown in FIG. 13 is composed of English and symbols. The language determination unit 941 selects the English TTS engine 932 as a selection engine that reads out the selected text. The beginning of the character string read out by the selection engine is “Hello!” And the end is “Hi!”.

  Selected text shown in Figure 14

Consists of English, Chinese and symbols. The language determination unit 941 selects the Chinese TTS engine 933 as a selection engine that reads out the selected text. The beginning of the string read by the selection engine is "Hello!"

It is.

[Details of selection engine decision processing]
A selection engine determination process executed by the language determination unit 941 will be described with reference to FIG. FIG. 15 is a flowchart showing the flow of the selection engine determination process. Here, the language determination unit 941 acquires the selected text from the temporary storage unit 61 in advance before executing the selection engine determination process. In the initial stage, the language setting is English and the variable n is 1.

  In the selection engine determination process, the language determination unit 941 first determines whether the nth character (or symbol) of the selected text is the last character of the selected text or the determination end symbol (S100). Here, the determination end symbol may be an external character or a line feed symbol.

  When the nth character of the selected text is the last character of the selected text or the determination end symbol (Yes in S100), the language determination unit 941 determines the selection engine based on the current language setting, and The nth character of the selected text or the character immediately before the nth character is determined as the end of the character string read out by the selection engine (S600). Then, the selection engine determination process ends.

  On the other hand, when the nth character of the selected text is not the last character of the selected text or the determination end symbol (No in S100), the language determining unit 941 determines that the nth character of the selected text is an alphabet or a symbol (determination). It is determined whether or not it is a symbol other than the end symbol (S200). When the nth character of the selected text is an alphabet or a symbol (Yes in S200), the language determination unit 941 adds 1 to the variable n (S500). Then, the selection engine determination process returns to the process of step S100 for determining whether or not the next character is the last character or the end determination character of the selected text.

  On the other hand, when the nth character of the selected text is not an alphabet or a symbol (No in S200), the language determination unit 941 determines whether or not the nth character is Japanese (S300). When the n-th character is Japanese (Yes in S300), the language determination unit 941 determines the language (A: Japanese, B, English, C: other languages) at that time. Is determined (S310).

  If the language setting at step S310 is Japanese (A in S310), the language determination unit 941 adds 1 to the variable n (S500). Then, the selection engine determination process returns to the process of step S100 for determining whether or not the next character is the last character or the end determination character of the selected text.

  If the language setting at step S310 is English (B in S310), the language determination unit 941 changes the language setting to Japanese (S320). Thereafter, the language determination unit 941 adds 1 to the variable n (S500). Then, the selection engine determination process returns to the process of step S100 for determining whether or not the next character is the last character or the end determination character of the selected text.

  If the language setting at the time of step S310 is a language other than Japanese and English (C in S310), the selection engine is determined based on the language setting at that time, and the nth character of the selected text The immediately preceding character is determined as the end of the character string read out by the selection engine (S600). Then, the selection engine determination process ends.

  In step S300, when the nth character is not Japanese (No in S300), the language determination unit 941 determines whether or not the nth character is Chinese (S400). When the n-th character is Chinese (Yes in S400), the language determination unit 941 determines the language setting (A: Chinese, B, English, C: other languages) at the time of step S400. It is determined whether there is (S410).

  When the language setting at the time of step S410 is Chinese (A in S410), the language determination unit 941 adds 1 to the variable n (S500). Then, the selection engine determination process returns to the process of step S100 for determining whether or not the next character is the last character or the end determination character of the selected text.

  If the language setting at the time of step S410 is English (B in S410), the language determination unit 941 changes the language setting to Chinese (S420). Thereafter, the language determination unit 941 adds 1 to the variable n (S500). Then, the selection engine determination process returns to the process of step S100 for determining whether or not the next character is the last character or the end determination character of the selected text.

  If the language setting at the time of step S410 is a language other than Chinese and English (C in S410), the selection engine is determined based on the language setting at that time and the nth character of the selected text The immediately preceding character is determined as the end of the character string read out by the selection engine (S600). Then, the selection engine determination process ends.

  Note that the method by which the language determination unit 941 determines the selection engine is not limited to the method described above. For example, the language determination unit 941 may select one TTS from one or more TTS engines that can read out a character string more than a predetermined reference among TTS engines that can read out the first character of the selected text. May be selected. For example, the predetermined criterion may be 70% of the number of character strings included in the selected text. In the case of this configuration, the language determination unit 941 selects any one of the TTS engines that read out 70% or more of the character string.

  Alternatively, the language determination unit 941 can generate more character string audio data from the character string after the focused character in the TTS engine corresponding to the focused character (for example, the first character that can be read out) in the selected text. A TTS engine that can generate Here, the characters that can be read out are not characters that cannot be read out such as symbols, but characters that can be read out such as alphabets and kanji.

[Embodiment 4]
Each control block (in particular, the kanji extraction unit 42, the reading kana extraction units 43 and 83, and the kanji kana replacement unit 44) included in the information processing apparatuses 1 and 2 is a logic circuit (hardware) formed in an integrated circuit (IC chip) or the like. Hardware), or software using a CPU (Central Processing Unit).

  In the latter case, the information processing apparatuses 1 and 2 include a CPU that executes instructions of a program that is software that realizes each function, and a ROM (Read Only) in which the program and various data are recorded so as to be readable by a computer (or CPU). Memory) or a storage device (these are referred to as “recording media”), a RAM (Random Access Memory) for expanding the program, and the like. And the objective of this invention is achieved when a computer (or CPU) reads the said program from the said recording medium and runs it. As the recording medium, a “non-temporary tangible medium” such as a tape, a disk, a card, a semiconductor memory, a programmable logic circuit, or the like can be used. The program may be supplied to the computer via an arbitrary transmission medium (such as a communication network or a broadcast wave) that can transmit the program. The present invention can also be realized in the form of a data signal embedded in a carrier wave in which the program is embodied by electronic transmission.

[Summary]
The text-to-speech device (information processing devices 1 and 2) according to aspect 1 of the present invention converts the text data to be displayed on the display unit (10) into voice data, and outputs the converted voice data, thereby A text-to-speech device that reads out text data, and a plurality of speech data generation units (Japanese TTS engine 931, English TTS engine 932, Chinese TTS engine 933) that respectively generate speech data in one or more languages, and a selection One voice that can determine the language type of the character string within the range of the text data and generate voice data of a plurality of types of character strings included in the text data from the plurality of voice data generation units And a selection unit (language determination unit 941) for selecting a data generation unit.

  According to the above configuration, based on the language type of the character string included in the range of the selected text data, the language included in the text data is selected from the plurality of speech data generation units. One voice data generation unit capable of generating voice data is selected.

  Therefore, even in the case where character strings of a plurality of languages are included in the range of the selected text data, a voice data generation unit suitable for reading out a plurality of types of text data is used, Sound data of a character string is generated. Accordingly, text data including character strings of a plurality of types of languages can be read out smoothly.

  In the text-to-speech apparatus (information processing apparatuses 1 and 2) according to aspect 2 of the present invention, in the aspect 1, the selection unit (language determination unit 941) is a voice of a character string of interest in the selected text data. A voice data generator (Japanese TTS engine 931, English TTS engine 932, Chinese TTS engine 933) that can generate voice data of more character strings in the selected text data from the voice data generator that can generate data. ) May be selected.

  According to said structure, the audio | voice data audio | voice which can produce | generate the audio | voice data of the character string of interest in the selected said text data, and can produce | generate the audio | voice data of more character strings in the selected said text data A data generator can be selected.

  In the above configuration, the character string of interest may be the first character that can generate voice data when viewed from the beginning of the selected text data.

  In this case, as viewed from the beginning of the selected text data, the voice data generation unit corresponding to the first character (for example, alphabet or kanji) that can generate voice data, that is, the first character voice that can generate voice data An audio data generation unit capable of generating data is selected.

  In addition, when there are a plurality of voice data generation units corresponding to the first character capable of generating voice data when viewed from the beginning, among the voice data generation units, more character strings in the character strings after the first character It is possible to select an audio data generation unit capable of generating the audio data.

  The text-to-speech apparatus (information processing apparatus 1 or 2) according to aspect 3 of the present invention is the above-described aspect 1 or 2, wherein the plurality of voice data generation units (Japanese TTS engine 931, English TTS engine 932, Chinese TTS engine) 933) includes a first audio data generation unit that generates audio data of the first language and the second language, and a second audio data generation unit that generates audio data of the first language and the third language. And the selection unit selects the first speech data generation unit when the first language is included in the range of the selected text data, and (i) when the second language is further included. In addition, (ii) when the third language is further included, the second audio data generation unit may be selected.

  According to the above configuration, the first voice data generation unit generates the voice data of the character strings of the first language and the second language, while the second voice data generation unit generates the first language and the second language. The voice data of the character string of the 3 languages is generated.

  Therefore, text data including the first language and the second language, and text data including the first language and the third language can be read out smoothly.

  The text-to-speech device (information processing devices 1 and 2) according to aspect 4 of the present invention is the selection unit (language determination unit 941) within the range of the selected text data in any of the above aspects 1 to 3. If there is a character string that cannot generate speech data in the speech data generation unit (Japanese TTS engine 931, English TTS engine 932, Chinese TTS engine 933) selected by the above, the selection unit You may further select the said audio | voice data production | generation part which can produce | generate the audio | voice data of a character string.

  According to the above configuration, even if there is a character string that cannot be generated in the voice data generation unit selected by the selection unit within the range of the selected text data, A sound data generation unit that generates sound data of the column is newly selected. Therefore, a wider range of character strings is read out in the selected text data.

  For example, when the selected text data includes English (first language), Japanese (second language), and Chinese (third language) character strings, the character string of each language is read out. A character string of each language is read out by the adapted voice data generation unit. Accordingly, it is possible to smoothly read a text including a character string of three or more languages.

  The text-to-speech device (information processing devices 1 and 2) according to aspect 5 of the present invention converts the text data to be displayed on the display unit (10) into voice data, and outputs the converted voice data, thereby A text-to-speech device that reads out text data and associates the text data including a kanji character string and a kana character string with the kanji character string within a range of the selected text data when the text data is displayed on the display unit. The extracted kana character string is extracted by the kana character string extracting unit (reading kana extraction unit 43, 83) that extracts the kana character string from the text data as a reading kana, and the kana character string extracting unit instead of the kanji character string. A kanji character string voice data generation unit (TTS) that generates voice data for reading out the kanji character string based on the kana character string. Engine 30 and the kanji kana replacement section 44), and a.

  According to the above configuration, when a kanji character string within a range in the selected text data is read out, first, a kana character string associated with the kanji character string is extracted as a reading kana from the text data. Is done. Here, the kana character extraction unit may extract, for example, a hiragana or katakana enclosed in parentheses immediately after the kanji character string as a kana character string designated as a reading kana of the kanji character string. .

  Further, the extracted kana character string is converted into voice data. The kanji character string is read out based on the voice data converted from the kana character string. In other words, the kanji character string is read out by the reading indicated by the kana character string.

  As a result, even when there are multiple readings in the kanji character string in the text data, the kanji character string is read more correctly (ie, the content creator intends to read the kanji character string without referring to other than the text data). Reading). As a result, it is possible to improve the reading accuracy of text including a kanji character string.

  The text-to-speech apparatus (information processing apparatus 1) according to aspect 6 of the present invention is the above-described aspect 5, wherein the kana character string extraction unit (reading kana extraction unit 43) is a predetermined first character that exists immediately after the kanji character string. A kana character string surrounded by one symbol and a predetermined second symbol existing after the first symbol may be extracted.

  According to the above configuration, the kana character string surrounded by the first symbol and the second symbol existing immediately after the kanji character string is extracted as a kana character string associated with the kanji character string. For example, when the first symbol and the second symbol are parenthesis opening and closing parenthesis “()” and the text is “Ogami”, the parenthesis immediately after the kanji character string “Okami” A kana character string “Oomiwa” surrounded by “()” is extracted as a kana character string associated with the kanji character string. Then, instead of the kanji character string “Okami”, the kana character string “Oomiwa” is read out.

  Here, the kana character string enclosed in parentheses immediately after the kanji character string is highly likely to be a reading kana of the kanji character string. For example, in the above-mentioned text “Ogami”, the kana character string “Oomiwa” is the reading kana of the kanji character string “Okami”. In this example, a kanji character string having a plurality of readings “Okami” can be read out as “Omiwa” as intended by the creator of the content, not “Okami” or “Dajin”. In addition, even if the text is other than this example, if the kana character string enclosed in parentheses immediately after the kanji character string is text that is a reading kana of the kanji character string, the kanji character string Can be read correctly. The predetermined first and second symbols may be, for example, “<>” or “<< >>” in addition to the parentheses opening and closing “()”. Alternatively, the first symbol and the second symbol may be the same symbol, or may be symbols unrelated to each other (for example, $ and #).

  In the text-to-speech device (information processing apparatus 1) according to aspect 7 of the present invention, in the above-described aspect 5 or 6, the text data has tag information that associates kanji in the text data with the reading kana. The kana character string extraction unit (reading kana extraction unit 43) may extract the reading kana of the kanji character string in the text data with reference to the tag information.

  According to the above configuration, based on the tag information, the reading kana of the kanji character string in the text data is extracted and read out instead of the kanji character string. Thereby, the text-to-speech device can read out the kanji character string in the text data with correct reading.

  In the text-to-speech device (information processing device 2) according to aspect 8 of the present invention, in the aspect 5, the text data includes ruby information indicating ruby attached to the kanji character string, and the kana character The column extraction unit (reading kana extraction unit 83) may extract the ruby of the kanji character string as the kana character string associated with the kanji character string with reference to the ruby information.

  According to the above configuration, the ruby of the kanji character string in the text data is extracted based on the ruby information included in the text data. As a result, the kanji in the text data can be read out with the reading intended by the creator of the content. This is because ruby information is often created by a content creator.

  The text-to-speech device (information processing devices 1 and 2) according to aspect 9 of the present invention is the kanji character string extraction unit (kanji character string) that extracts the kanji character string from the text data in any of the above aspects 5 to 8. The kana character string extraction unit (reading kana extraction units 43 and 83) extracts the kana character string associated with the kanji character string extracted by the kanji character string extraction unit. May be.

  According to the above configuration, the kana character string associated with the kanji character string in the text data is read out instead of the kanji character string extracted by the kanji character string extraction unit. This configuration can be applied, for example, when the text-to-speech device reads text data. When reading out text data, the text-to-speech device reads out character strings in the text data in order from the beginning. During the reading, if the character string to be read out next is a kanji character string, the kanji character string extraction unit extracts the kanji character string. The text-to-speech device reads the kana character of the kanji character string instead of the kanji character string. As a result, the text-to-speech device reads the text data with correct reading (that is, the reading intended by the content creator) even if there is a kanji character string having a plurality of readings in the text data. can do.

  The text-to-speech device (information processing devices 1 and 2) according to aspect 10 of the present invention further includes the display unit in aspect 5, and the kana character string extraction unit (reading kana extraction units 43 and 83) The kana character string associated with the kanji character string selected by the user may be extracted from the text data.

  According to said structure, the kanji character string in a text is selected by the user. Then, instead of the kanji character string selected by the user, the kana character string associated with the kanji character string in the text data is read out. As described above, when the kana character string associated with the kanji character string is extracted as the kana character string, the kana character string reading kana is used instead of the kanji character string. Read aloud.

  Thereby, when the user selects a kanji character string desired to be read out, the text-to-speech device can read out the kanji character string with correct reading even if the kanji character string has a plurality of readings. .

  A control method for a text-to-speech device according to an aspect 11 of the present invention is a text-to-speech device that reads out the text data by converting the text data to be displayed on the display unit into voice data and outputting the converted voice data. The kana character associated with the kanji character string within the range of the selected text data when the text data including the kanji character string and the kana character string is displayed on the display unit. A kana character string extraction step for extracting a string as a reading kana from the text data; and, instead of the kanji character string, the kanji character string is converted based on the kana character string extracted in the kana character string extraction step. A voice data generation step for generating voice data to be read out.

  According to said structure, there can exist an effect similar to the text-to-speech apparatus which concerns on aspect 5. FIG.

  The information processing device (1, 2) according to each aspect of the present invention may be realized by a computer. In this case, the text-to-speech device is operated by causing the computer to operate as each unit included in the text-to-speech device. A text-to-speech control program implemented by a computer and a computer-readable recording medium on which the control program is recorded also fall within the scope of the present invention.

  The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope shown in the claims, and embodiments obtained by appropriately combining technical means disclosed in different embodiments. Is also included in the technical scope of the present invention. Furthermore, a new technical feature can be formed by combining the technical means disclosed in each embodiment.

  The present invention can be suitably used for various text-to-speech devices that read text, such as an electronic dictionary, an electronic book terminal, and a tablet terminal.

1, 2 Information processing device (text-to-speech device)
30 TTS engine (kanji character string voice data generator)
42 Kanji extraction unit (Kanji character string extraction unit)
43, 83 Reading Kana extraction part (kana character string extraction part)
44 Kanji kana replacement part (kanji character string voice data generation part)
931 Japanese TTS engine (voice data generator, first voice data generator)
932 English TTS engine (voice data generator)
933 Chinese TTS engine (voice data generator, second voice data generator)
941 Language determination unit (selection unit)

Claims (3)

A text-to-speech device that reads out the text data by converting the text data to be displayed on the display unit into voice data and outputting the converted voice data,
A first audio data generation unit that generates audio data of a first language and a second language, and a second audio data generation unit that generates audio data of a first language and a third language, A plurality of voice data generation units that respectively generate voice data in the above languages ;
The language type of the character string within the range of the selected text data can be determined, and the voice data of the character strings of a plurality of types of languages included in the text data can be generated from the plurality of voice data generation units. A selection unit for selecting an audio data generation unit ,
When the first language is included in the range of the selected text data, the selection unit selects (i) the first speech data generation unit when the second language is further included, and (ii) A text-to-speech device that selects the second voice data generation unit when a third language is further included .   The selection unit can generate voice data of more character strings in the selected text data from a voice data generation unit that can generate voice data of a character string of interest in the selected text data. The text-to-speech apparatus according to claim 1, wherein a generation unit is selected. When there is a character string that cannot be generated in the voice data generation unit selected by the selection unit within the range of the selected text data, the selection unit displays the voice of the character string. The text-to-speech apparatus according to claim 1 or 2 , further selecting another voice data generation unit capable of generating data.

Images