JP2001318915A - Font conversion device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To recognize a sentence spoken by a speaker as voice and to convert the font of a text based on voice information indicating the feature of an inputted voice. SOLUTION: When a speaker speaks a message, voice data are inputted from a voice input part 101 to a voice recognition part 105. The voice recognition part 105 recognizes the voice and converts the message into a text. A power information extraction part 106 extracts voice power from the voice data and applies the extracted power to a font conversion control part 102. On the other hand, font information corresponding to each voice power is stored in a font table 1001. A font conversion control part 102 extracts required font information from the font table 1001 through a font table reference part 104 on the basis of the obtained voice power, converts the font of the input text and outputs the converted font on a display part 103.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a font conversion apparatus for performing a font conversion on an input voice or text based on voice information indicating characteristics of the input voice or reliability information of the voice input text. .

[0002]

2. Description of the Related Art Some document processing apparatuses such as a Japanese word processor can input voice. Such a document processing device recognizes the input voice and converts the text to display the text on a display or the like.

[0003] For example, the technology disclosed in Japanese Patent Application Laid-Open No. Sho 60-129795 discloses a speaker recognition technology for recognizing voice and determining whether the voice is male or female. . In such a technique, a gender-specific standard pattern is used to recognize input speech of an unspecified speaker, and the gender of the speaker is determined from the recognition result.

[0004] In the technique disclosed in Japanese Patent Application Laid-Open No. 05-323990, a speaker identification technique is described. In such a technique, a model is created and registered for each speaker, and the speaker that uttered the voice is recognized by recognizing the voice using the model.

[0005]

However, in the conventional document processing apparatus, when changing the font of a document, it is necessary to instruct a font change from a predetermined menu using an input means such as a mouse. Further, the conventional document processing apparatus capable of voice input only displays a document whose speech has been recognized, and the certainty of recognition of the recognized document has not been reflected in the text.

The present invention has been made in view of such a conventional problem. The inventions according to claims 1 to 10 enable a text font to be changed by voice input, and the invention according to claim 11 SUMMARY OF THE INVENTION It is an object of the present invention to realize a font conversion device which can reflect the certainty of recognition of a character input by voice in a font of a text.

[0007]

According to a first aspect of the present invention, a speech uttered by a speaker is recognized and converted into text data, and the text is converted to text data based on speech information included in the input speech of the speaker. A font conversion device for converting a font of data, wherein a voice uttered by a speaker is input, and a voice input unit for outputting voice data; A voice recognition unit that converts data into data, a voice information extraction unit that extracts voice information indicating voice characteristics of a speaker from voice data input by the voice input unit, and discriminates the voice information into a plurality of classes or attributes. A font table for storing text font information corresponding to each class or attribute; and font information corresponding to input voice information with reference to the font table. When the voice information of the speaker is provided from the font table reference unit to be obtained and the voice information extraction unit and the text data of the speaker is provided from the voice recognition unit, the voice information is provided to the font table reference unit. A font conversion control unit that performs font conversion on the text data using font information obtained from the font table, and a display unit that displays text converted by the font conversion control unit. Is what you do.

According to a second aspect of the present invention, in the font conversion apparatus according to the first aspect, the voice information extracting unit uses the voice power of a speaker as the voice information from the voice data input from the voice input unit. It is a power information extracting unit to be extracted.

According to a third aspect of the present invention, in the font conversion apparatus according to the first aspect, the voice information extracting unit extracts a voice feature value used for speaker identification from the voice data input by the voice input unit. A speaker identification feature extraction unit that extracts as the voice information, a speaker identification label acquisition unit that inputs a speaker identification label corresponding to each speaker, and a voice extracted from the speaker identification feature extraction unit A speaker feature learning unit that creates a template used for speaker identification using the feature amount and the speaker identification label input from the speaker identification label acquisition unit, and an extraction from the speaker identification feature extraction unit A speaker identification unit that identifies a speaker using the speech feature amount to be output and the template created by the speaker feature learning unit, and outputs a speaker identification label of the speaker who uttered the speech; From the user identification label acquisition unit Was by using the speaker identification label and font information specified by each speaker, it is characterized in that it has a font table creating unit for creating a font table, a.

According to a fourth aspect of the present invention, in the font conversion apparatus according to the first aspect, the voice information extracting unit is configured to convert a voice feature value used for identifying a speaker's emotion from voice data input by the voice input unit. And an emotion identification unit that identifies a speaker's emotion by using the audio feature amount extracted from the emotion identification feature extraction unit. It is assumed that.

According to a fifth aspect of the present invention, in the font conversion apparatus according to the first aspect, the voice information extracting unit uses a voice feature value used for gender identification of a speaker based on voice data input by the voice input unit. And a gender identification unit that identifies the gender of the speaker by using the audio feature extracted from the gender identification feature extraction unit. It is assumed that.

According to a sixth aspect of the present invention, in the font conversion apparatus according to the first aspect, the voice information extracting unit is configured to determine, based on the voice data input by the voice input unit, a voice feature value used for identifying a speaker's age. As an audio information, and an age identification unit that identifies the age of the speaker by using an audio feature extracted from the age identification feature extraction unit. It is assumed that.

According to the invention of claim 7 of the present application, a font table for storing font information corresponding to each class of voice feature is newly created, and a voice uttered by a speaker is recognized and converted into text data. A font conversion device for converting a font of the text data based on voice information included in an input voice of a speaker using the font table, inputting a voice uttered by the speaker and outputting voice data A voice input unit, a voice recognition unit that recognizes voice data input from the voice input unit and converts the voice data into text data, and indicates a voice characteristic of a speaker from the voice data input by the voice input unit. A voice information extraction unit for extracting voice information, a font designation input unit for inputting font information, and discriminating the voice information into a plurality of classes, corresponding to each class Using a font table that stores font information of the extracted text and font information input by the font designation input unit and audio information extracted by the audio information extraction unit, A font table creation unit for newly storing a font table; a font table reference unit for referring to the font table to obtain font information corresponding to input audio information; and a speaker voice from the audio information extraction unit. Information, and when the speaker's text data is provided from the voice recognition unit, the voice information is provided to the font table reference unit, and the text data is converted to a font using the font information obtained from the font table. A font conversion control unit to be converted, and a text converted by the font conversion control unit. Is characterized in that it comprises a display unit for displaying the door, the.

According to an eighth aspect of the present invention, the voice uttered by the speaker is recognized and converted into text data, and the font of the text data is converted based on voice information included in the input voice of the speaker. A font input device for inputting a voice uttered by a speaker and outputting voice data when the font change of the text is executed, and the changed font is used for subsequent font conversion. A voice recognition unit that recognizes voice data input from the voice input unit and converts the voice data into text data; and extracts voice information indicating a voice characteristic of a speaker from the voice data input by the voice input unit. A speech information extracting unit, and a font table for discriminating the speech information into a plurality of classes and storing text font information corresponding to each class , Above with reference to the font table, a font table reference unit for acquiring font information corresponding to the input voice information, text information including a display position of the text and the text which speaker to change fonts,
And a font change input unit for inputting font information after the font change, and the voice information extracting unit, when the speaker's voice information is provided, and when the speaker's text data is provided from the voice recognizing unit, the font table. The audio information is provided to the reference unit, and the text is converted into a font using the font information obtained from the font table, and the audio information used when the text input from the font change input unit is subjected to the font conversion is output. A font conversion control unit that changes the font of the text to be changed input from the font change input unit with the font information input from the font change input unit; and audio information output from the font conversion control unit. The font table based on font information input from the font change input unit; And font learning unit for changing the data, is characterized in that it comprises a display unit for displaying the converted text with the font conversion control unit.

According to a ninth aspect of the present invention, a speaker speaks text data that has already been input, thereby extracting voice information of the speaker with respect to each input text data, and extracting the voice information of the input text data. A font conversion device for converting a font, comprising: a voice input unit for inputting voice uttered by a speaker and outputting voice data; and a voice characteristic of the speaker based on voice data input by the voice input unit. A voice information extraction unit for extracting voice information to be indicated, a font table for discriminating the voice information into a plurality of classes, and storing text font information corresponding to each class, and an input voice with reference to the font table. Font table reference section for obtaining font information corresponding to information, and text input for inputting text information including text and text display position And a text-to-speech designation display unit for designating a part of the text input by the text input unit, instructing a speaker to speak the designated text, and outputting text information of the designated text, When the voice information of the speaker is given from the information extracting unit and the text information is given from the reading designation display unit,
The voice information is given to the font table reference unit, and the font information obtained from the font table is used. And a display unit for displaying the text converted by the font conversion control unit while displaying the converted text.

According to a tenth aspect of the present invention, a speaker speaks text data that has already been input, thereby extracting speaker voice information for each input text data.
A font conversion device for converting a font of the input text data, wherein a voice uttered by a speaker is input, and a voice input unit that outputs voice data, and a voice data input from the voice input unit. A voice recognition unit that performs voice recognition and converts the voice information into text data; a voice information extraction unit that extracts voice information indicating voice characteristics of a speaker from voice data input by the voice input unit; A font table for storing font information of text corresponding to each class, a font table reference unit for referring to the font table and obtaining font information corresponding to input voice information, and a text and a text. A text input unit for inputting text information including a display position, a text recognized by the voice recognition unit, and the text input It compares the input text from,
When a position information detecting unit that outputs the text information having the same comparison result is provided with the speaker's voice information from the voice information extracting unit, and when the speaker's uttered text information is provided from the position information detecting unit. A font conversion control unit that provides the audio information to the font table reference unit, converts the text output from the position information detection unit into a font using the font information acquired from the font table, and the text input unit. A display unit for displaying the input text and displaying the text converted by the font conversion control unit.

An invention according to claim 11 of the present application is a font conversion apparatus for recognizing voice uttered by a speaker and converting the voice into text data, and converting the font of the text data based on the certainty of voice recognition. A voice input unit for inputting voice uttered by a speaker and outputting voice data; and converting voice data input from the voice input unit into text data by voice recognition. A speech recognition unit that outputs the likelihood of the likelihood information as likelihood information, the likelihood information is discriminated into a plurality of classes, and a font table that stores font information corresponding to each class, and with reference to the font table, A font table reference unit for obtaining font information of text corresponding to the input likelihood information; and a speaker text data and text from the speech recognition unit. When the likelihood information of the font table is given, the likelihood information is given to the font table reference unit, and a font conversion control unit that performs font conversion of the text using font information acquired from the font table; A display unit for displaying the text converted by the font conversion control unit.

[0018]

(Embodiment 1) Hereinafter, a font conversion apparatus according to Embodiment 1 of the present invention will be described with reference to the drawings. FIG. 1 is a configuration diagram of a font conversion apparatus according to Embodiment 1 of the present invention. This font conversion device includes a voice input unit 101, a font conversion control unit 10
2, a display unit 103, a font table reference unit 104, a speech recognition unit 105, a power information extraction unit 106, and a font table 1001.

The voice input unit 101 takes in a message (voice) uttered by a speaker and outputs voice data. The power information extraction unit 106 is a voice input unit 1
When audio data is input from 01, the audio information extraction unit extracts power information as audio information of the input audio data. When voice data is input, the voice recognition unit 105 performs voice recognition on the voice data, and outputs text data as a recognition result in units of words, phrases, or sentences. The conversion unit of this is determined by the speaker's utterance state (break between words, presence or absence of pause).

The font table 1001 discriminates audio power information into a plurality of classes or attributes and stores font information corresponding to each class or attribute. FIG. 2 shows an example of the font table 1001. In FIG. 2, the audio power is less than 0 dB, 0 to 10 dB.
B,... A plurality of classes are set with a 10 dB width such as 40 to 50 dB. As the font setting value as font information, a combination of a font name, style, size, character decoration, color, and the like is set for each audio power.

The font table reference unit 104 shown in FIG.
When power information is input as input audio information from the font conversion control unit 102, font information corresponding to the input power information is acquired by referring to the font table 1001, and the font information is transmitted to the font conversion control unit 102. Is to give.

The font conversion control unit 102 receives the power information from the power information extraction unit 106 and
When the speaker's text data is input from the input unit 05, the power information is output to the font table reference unit 104, and based on the font information output from the font table reference unit 104, the text data input from the speech recognition unit 105 is output. Is used for font conversion. The display unit 103 displays text (sentence) of a standard font before font conversion, and displays font-converted text. Note that the information of the font table 1001 may be temporarily read into the memory, and the font conversion control unit 102 may refer to the information.

An operation example of the thus configured font conversion apparatus according to the present embodiment will be described. In FIG. 1, a speech information input unit 101 outputs a power information extraction unit 106
As an example, it is assumed that voice data “Tomorrow will be fine.” Is input to the voice recognition unit 105 by the speaker. In this case, the speaker gives 45 tomorrow
Suppose that the user uttered the speech at the size of dB. The voice data of “Tomorrow is” is transmitted from the voice input unit 101 to the power information extraction unit 1.
06 and the speech recognition unit 105, the power information extraction unit 106 extracts the power information of the speech data and outputs “4”.
5 dB ". This power information is provided to the font table reference unit 104 via the font conversion control unit 102.

On the other hand, the voice recognition unit 105 performs voice recognition on the input voice data, and provides the text data of the recognition result to the font conversion control unit 102. In the above example, the voice recognition unit 105 performs voice recognition on the voice data “Tomorrow is,” converts the voice data into the text “Tomorrow is,” and outputs the text data to the font conversion control unit 102.

The font table reference unit 104 refers to the font table 1001 in FIG.
Font information "(Font) Gothic,
(Style) standard, (size) 12, (no character decoration), (color) black ”are obtained, and this font information is provided to the font conversion control unit 102.

The font conversion control unit 102 converts the text "tomorrow is" input from the speech recognition unit 105 based on the above-mentioned font information provided from the font table reference unit 104, and performs the font conversion shown in FIG. ) Has a font such as “(Font) Gothic, (Style) Standard, (Size) 12, (Character decoration)
None, (color) black ”is displayed on the display unit 103.

Next, it is assumed that the speaker utters "sunny" with a magnitude of 4 dB, and subsequently utters "will" with a magnitude of 41 dB. The font table reference unit 104 refers to the font table 1001 again, and outputs the power information “4d
Font information "(Font) Gothic,
(Style) bold, (size) 24, (no character decoration), (color) black ”, and provides this font information to the font conversion control unit 102. Next, the font table reference unit 104 outputs the power information “41 dB” of “Would.”
Font information corresponding to "(Font) Gothic, (Style) Standard, (Size) 12, (No character decoration),
(Color) black ”and gives the font information to the font conversion control unit 102. As a result, the display unit 1 displays the font "Tomorrow will be fine" as shown in FIG.
Displayed as 03.

In this way, the font of the character input by voice can be converted and displayed on the screen according to the strength of the utterance of the speaker. For example, if the power is equal to or higher than a certain power, the normal font and the emphasized font can be selectively used depending on the strength of the voice by using the emphasized font.

(Embodiment 2) Next, a font conversion apparatus according to Embodiment 2 of the present invention will be described with reference to the drawings. In the above-described first embodiment, the fixed font table 1001 is referred to. However, the font conversion apparatus according to the present embodiment creates a new font table 1001 before performing voice input, and thereafter, the speaker performs voice input. Then, the power information of the input voice data is extracted, and the text data of the voice recognition result is output in a newly created font.

FIG. 4 is a block diagram of the font converter according to the present embodiment. Here, it is assumed that the same blocks as those in the first embodiment perform the same operations, and the same reference numerals are given, and detailed descriptions thereof are omitted.

The font conversion apparatus according to the present embodiment includes a voice input unit 101, a font conversion control unit 102, a display unit 10
3. Font table reference unit 104, voice recognition unit 10
5. Power information extraction unit 106, font table 100
In addition to the above, a font designation input unit 107 and a font table creation unit 108 are included.

The font designation input unit 107 is composed of an input device such as a keyboard, and is used by a user to input font information. When power information, which is audio information, is input from the power information extraction unit 106 and font information is input from the font designation input unit 107, the font table creation unit 108 generates a font table 1001 based on the power information and the font information. Is newly created.

An example of the operation of the font converting apparatus according to the present embodiment having the above-described configuration will be described. In FIG. 4, a speaker utters a test text before voice input of an actual text, and gives test voice data from a voice input unit 101 to a power information extraction unit 106. As a specific example, a speaker utters a test sentence with strong and weak. At this time, the strong utterance has a size of 10 dB,
Weak utterances are 30 dB in magnitude. These voice data are transmitted to the power information extraction unit 106 via the voice input unit 101.
Is input to

Power information extracting section 106 extracts power information of the audio data input from audio input section 101. The power information extracted here is output to the font table creation unit 108. In the above example, the power information extraction unit 106 recognizes the power information of the voice input from the voice input unit 101 as 10 dB and 30 dB, and outputs the power information to the font table creation unit 108. That is, it is determined that the voice power of the current speaker fluctuates around the range of 10 dB to 30 dB.
Three classes of less than 10 dB, 10 dB or more and less than 30 dB, and 30 dB or more are set. The state of the font table 1001 at this point is as shown in FIG. The width of the class can be set arbitrarily.

Next, the speaker inputs font information from the font designation input unit 107 to the font table creation unit 108. In the above example, font information such as “(Font) Gothic, (Style) bold, (Size) 32, (Character decoration) embossed, (Color) red” is set for a class of less than 10 dB. For classes of 10 dB or more and less than 30 dB, “(Font) Gothic, (Style) bold, (Size) 24, (No character decoration), (Color)
Set black font information. Furthermore, for the class of 30 dB or more, "(Font) Gothic, (Style) Standard, (Size) 12, (No character decoration), (Color)
Set black font information. Thus, FIG.
A font table 1001 as shown in FIG.

After the font table 1001 is created, the speaker inputs the actual text by voice. In this case, since the font table 1001 has been created, the same operation as in the first embodiment is performed. That is, the font table reference unit 104 stores the newly created font table 1
Referring to 001, font information is obtained based on the power information, and is provided to the font conversion control unit 102.

In this way, the font of the character input by voice can be converted into a value set in advance by the user and displayed on the screen according to the strength of the speaker's utterance. That is, by learning the voice power information of the individual, it is possible to absorb individual differences in voice volume.

(Embodiment 3) Next, a font conversion apparatus according to Embodiment 3 of the present invention will be described with reference to the drawings. After the font conversion apparatus of the second embodiment described above has created the font table 1001, it was not possible to change the font information of the font table 1001 during voice input. However, the font conversion apparatus according to the present embodiment is characterized in that a speaker can view a text displayed on display unit 103 and change the font of the text. Further, the font conversion apparatus according to the present embodiment is characterized in that the font change of the text after the font conversion is reflected in the font table reference unit 104, and the font table reference function can be updated.

FIG. 6 is a block diagram of the font converter according to the present embodiment. The same reference numerals are given to the same blocks as in the second embodiment, and detailed description thereof will be omitted. The font conversion apparatus according to the present embodiment includes a voice input unit 1
01, font conversion control unit 109, display unit 103, font table reference unit 104, voice recognition unit 105, power information extraction unit 106, font table 1001, font learning unit 110, font change input unit 111
It is comprised including.

The font change input unit 111 is composed of an input device such as a mouse and a keyboard. The text displayed on the display unit 103 is used to convert text desired by the speaker and text information including display position information of the text. This is for inputting or inputting font information after the change. The text information and font information entered here are
The font information is output to the font conversion control unit 109, and the font information is also output to the font learning unit 110.

When the font information after the change is input from the font change input unit 111 and the power information is further input from the font conversion control unit 109, the font learning unit 110 refers to the font table 1001 and executes the font conversion control unit. Using the power information input from 109 and the font information input from the font change input unit 111,
A part of the data of the font table 1001 is updated.

The font conversion control unit 109 has the functions of the font conversion control unit 102 of the first and second embodiments,
When text information and font information are input from the font change input unit 111, the font of the input text is converted using the font information, and the text whose font has been changed is output to the display unit 103. . In addition, when text data is input from the font change input unit 111, the font conversion control unit 109 has a function of providing power information used for changing the font of the text data to the font learning unit 110.

An example of the operation of the thus configured font conversion apparatus according to the present embodiment will be described. In FIG. 6, when speech data generated by a speaker's utterance is given to the speech recognition unit 105, the speech is recognized and text data as a recognition result is output. Power information extraction unit 10
The font conversion control unit 109 converts the font of the text based on the power information of the audio extracted in step 6, and causes the display unit 103 to display the font-converted text. The above operation is the same as in the second embodiment.

It is assumed that after the font-converted text is displayed on the display unit 103, the speaker inputs the changed font information and the changed text information while looking at the display contents. In this case, the text information to be changed and the font information after the change are input to the font conversion control unit 109 via the font change input unit 111, and the font information after the change is also input to the font learning unit 110.

As an example, as shown in FIG. 7A, the text "Tomorrow will be fine."
It is assumed that it is displayed as 3. The font table 1001 referred to when the text of FIG. 7A is converted into a font has the contents shown in FIG. 8A. Since the text “sunny” in FIG. 7A was uttered at a size of 12 dB, the font table 10 in FIG.
At 01, font information in a class in which the magnitude of audio power is 10 dB or more and less than 30 dB is referred to. Next, in order for the speaker to change the text “sunny” in FIG. 7A from “bold” to “bold italic”, the text information of “sunny” and the changed font information “bold” are input from the font change input unit 111. "Italic". The text information “sunny” and the font information “bold italic” after the change are output to the font conversion control unit 109. The changed font information “bold italic” is also output to the font learning unit 110.

The font conversion control unit 109 outputs to the font learning unit 110 power information used when the text data input from the font change input unit 111 is subjected to font conversion. In the above example, the font conversion control unit 109 sends “fine”
Is entered. Then, the power information 12 dB used when performing font conversion on the text data of “sunny” is output to the font learning unit 110.

The font conversion control unit 109 changes the font of bold “clear”, which is the text input from the font change input unit 111, to bold italic “clear”. As a result, “clear” in bold italic as shown in FIG. Font learning unit 11
0 is the power information 12d from the font conversion control unit 109.
When B is input, the style item of the font information of the class of audio power of 10 dB or more and 30 dB is rewritten to the changed font information “bold italic” input from the font change input unit 111. As a result, the font table 1001 is updated as shown in FIG. 8B.

Next, suppose that the speaker utters, "The day after tomorrow will be rain." At this time, if the voice power of “rain”, which is a part of the utterance, is extracted as 20 dB, the text “rain” is font-converted using the font table 1001 of FIG. 8B, and is shown in FIG. 7C. Thus, the style of the text "Rain" is bold italic. As described above, in the present embodiment, the font change of “sunny” is reflected on the font conversion of “rain” that is input next by voice.

The font conversion apparatus according to the present embodiment is further provided with a font table creation section 108 according to the second embodiment, and the font designation input section 107 according to the second embodiment is replaced with the font change input section according to the third embodiment. The font table 1001 of the present embodiment may be newly created using the unit 111.

According to the font conversion apparatus of the present embodiment, a function of correcting the converted font at any time is realized, and by learning the contents of the change, the text can be converted to a font according to the speaker's preference. Can be converted.

(Embodiment 4) Next, a font conversion apparatus according to Embodiment 4 of the present invention will be described with reference to the drawings. In the above-described first to third embodiments, the apparatus has been described in which the text input by voice is font-converted and then the font-converted text is displayed on the display unit 103. The font conversion apparatus according to the present embodiment is characterized in that the font of already input text can be changed.

FIG. 9 is a configuration diagram of a font conversion apparatus according to the fourth embodiment of the present invention. The same blocks as those in the first embodiment are denoted by the same reference numerals, and detailed description thereof will be omitted. The font conversion apparatus according to the present embodiment includes a voice input unit 101,
In addition to the font conversion control unit 102, the display unit 103, the font table reference unit 104, the power information extraction unit 106, and the font table 1001, a text input unit 112,
It is configured to include a reading designation display unit 113.

The text input unit 112 is composed of an input device such as a keyboard, and allows a speaker to input text data.
It outputs text information including text data and a text display position. The text data input here is output to the display unit 103 and displayed, and the text information is output to the reading designation display unit 113.

Based on the text information input from the text input unit 112, the reading designation display unit 113 indicates the text to be read by the speaker such as for each character, for each word, for each sentence, etc. The font change order is automatically designated at a constant speed, and read-out text data (designated display text) is given to the display unit 103. The data of the designated display text is also output to the font conversion control unit 102.

An example of the operation of the thus configured font conversion apparatus according to the present embodiment will be described. As an example, the text "Tomorrow will be fine." Is input from the text input unit 112, and this text is shown in FIG.
Consider a case where the display unit 103 is displaying a standard font like 0 (a). When the above text information is given from the text input unit 112, the reading designation display unit 1
Reference numeral 13 generates a reading designation text, such as for each character, each word, or each sentence, and outputs it to the display unit 103. In addition, the text information of the designated reading-out text is also output to the font conversion control unit 102. In the above example, the text-to-speech designation display unit 113 first designates “Tomorrow is” as text to designate text-to-speech, and the display unit 103 displays the text underlined as shown in FIG.

The speaker looks at the text "Tomorrow is," which is designated and displayed on the display unit 103, and writes "Tomorrow is," 20 characters.
Suppose that the user uttered the speech at the size of dB. The audio input unit 101 supplies the audio data to the power information extraction unit 106. The power information extraction unit 106 detects that the power of the audio data “Tomorrow is” input from the audio input unit 101 is 20 dB. When this power information is input, the font conversion control unit 102 gives “20 dB” to the font table reference unit 104.

The font table reference unit 104 refers to the font table 1001 and uses the font conversion control unit 1
The font information corresponding to the power information input from 02 is output to the font conversion control unit 102. The font table reference unit 104 refers to, for example, a font table 1001 shown in FIG. Then, in the above example, a class including the audio power of 20 dB is searched, and font information corresponding to the audio power is read. 8A, “(Font) Gothic, (Style) bold, (Size) 24, (No character decoration), (Color)
The “black” font information is output to the font conversion control unit 102.

The font conversion control section 102 converts the text input from the reading designation display section 113 into a font based on the font information input from the font table reference section 104. The text after the font conversion is displayed on the display unit 103 based on the display position information input from the reading designation display unit 113. In the above example, the text “Tomorrow is” input from the reading designation display unit 113 is font-converted, and FIG.
“Tomorrow is” font-converted as shown in FIG.

Next, the reading-out designation display unit 113 causes the display unit 103 to designate and display the text next to the currently designated text after a predetermined time. In the above example, FIG.
As shown in (d), the text designated and displayed the previous time is “Tomorrow is,” so the next text “Sunny” is designated and displayed on the display unit 103.

Note that, when the speaker inputs a command for designating and displaying the next text into the reading designation display unit 113,
The text-to-speech designation display unit 113 may designate and display the next text. In this case, the font conversion interval can be adjusted to the speaker's pace.

Further, the font table creation unit 108 and the font designation input unit 107 according to the second embodiment are provided in the font conversion apparatus according to the present embodiment, and the font table 1001 according to the present embodiment is newly created. Good.

Further, the font change input unit 111 and the font learning unit 110 according to the third embodiment are provided in the font conversion device according to the present embodiment, and the third embodiment is replaced with the font conversion control unit 102 according to the third embodiment. The font table 1001 according to the present embodiment may be changed using the font conversion control unit 109 described above.

As described above, according to the font conversion apparatus of the present embodiment, a part of the input text is specified, and the specified text is spoken by the speaker, whereby the specified text is converted. Fonts can be freely changed according to audio power.

(Fifth Embodiment) Next, a font conversion apparatus according to a fifth embodiment of the present invention will be described with reference to the drawings. The above-described font conversion apparatus according to the fourth embodiment is an apparatus in which text whose font can be changed is specified by underlining or the like. The font conversion device according to the present embodiment
It is characterized in that a text to be converted can be automatically detected by recognizing a voice spoken by a speaker and comparing it with a sentence already input.

FIG. 11 is a block diagram of the font converter according to the present embodiment. The same blocks as those in the first and fourth embodiments are denoted by the same reference numerals, and detailed description thereof will be omitted. This font conversion device includes a voice input unit 101, a font conversion control unit 102, a display unit 103, a font table reference unit 104, a voice recognition unit 105, a power information extraction unit 106, a text input unit 112, a font table 1
001 and a position information detection unit 115.

The position information detecting section 115 refers to the text information input from the text input section 112 to compare and detect the same text as the text input from the voice recognizing section 105. It outputs text information including display position information and text data to the font conversion control unit 102.

An example of the operation of the font conversion apparatus according to the present embodiment thus configured will be described. Text information is provided from the text input unit 112 to the position information detection unit 115 by a user operation. The display unit 103 displays the text input from the text input unit 112. As an example, the text input unit 112
Then, text data "Tomorrow will be fine." Is input, and "Tomorrow will be fine." Is displayed in the standard font on the display unit 103 as shown in FIG. At this time, the text information “Will be fine tomorrow.” Is also input from the text input unit 112 to the position information detection unit 115.

Next, the speaker (user) utters the text displayed on the display unit 103. When a speaker's voice is input, the voice input unit 101 provides voice data to the power information extraction unit 106 and the voice recognition unit 105. In the above example, it is assumed that the speaker utters “Tomorrow is” with a magnitude of 20 dB. The power information extraction unit 106 includes a voice input unit 101
It is extracted that the power information of the audio data input from is 20 dB. This power information is output to the font conversion control unit 102.

On the other hand, the voice recognition unit 105 is
The voice data input from step 1 is recognized by voice and the text "Tomorrow is" is output. This text is provided to the position information detection unit 115. The position information detection unit 115 uses the text information input from the text input unit 112 and the text data input from the voice recognition unit 105 to write the same text data as the text data input from the voice recognition unit 105. And detects the display position of the text. The display position information and the text data are output to the font conversion control unit 102 as text information. In the above example, the text information “Tomorrow is fine.” Is input from the text input unit 112 to the position information detection unit 115. Also, the text “Tomorrow is” is input from the voice recognition unit 105 to the position information detection unit 115. The position information detection unit 115 inputs the text information of “Tomorrow will be fine.”
Compare with the text "Tomorrow is." Then, the position information detection unit 115 displays the text “Tomorrow is,”
In step 3, the position information indicating the text “Tomorrow is” is output to the font conversion control unit 102.

The font conversion control section 102 outputs the power information 20 dB input from the power information extraction section 106 to the font table reference section 104. The font table reference unit 104 refers to a font table 1001 as shown in FIG. Then, the font conversion control unit 10
Font information “(Font) Gothic, (Style) bold,
(Size) 24, (no character decoration), (color) black ”are obtained and output to the font conversion control unit 102.

The font conversion control unit 102 converts the text “tomorrow is” input from the position information detection unit 115 into “(font) gothic, (style) based on the font information input from the font table reference unit 104.
Bold, (size) 24, no (text decoration), (color) black "
Font conversion. The text "Tomorrow," which has been font-converted in this way, will use the standard font "Sunny."
At the same time, it is displayed on the display unit 103 as shown in FIG.

The font table creation unit 108 and the font designation input unit 107 according to the second embodiment are further provided in the font conversion apparatus according to the second embodiment.
1 may be newly created.

Further, the font change input unit 111 and the font learning unit 110 according to the third embodiment are further provided in the font conversion apparatus according to the present embodiment, and the font conversion control unit according to the third embodiment is used instead of the font conversion control unit 102. 109, the font table 1001 of the present embodiment may be changed.

As described above, according to the font conversion apparatus of the present embodiment, the input voice is recognized and compared with the already input text, so that the position information in the text of the text to be changed can be obtained. The gain function is added. This allows the speaker to read naturally.

(Embodiment 6) Next, a font conversion apparatus according to Embodiment 6 of the present invention will be described with reference to the drawings. In the font converters of the first, second, third, fourth, and fifth embodiments, audio power (power information) is used as audio information in order to convert text into a font. However, the font conversion apparatus according to the present embodiment is characterized in that a speaker who has uttered a text as voice information is identified, and font conversion is performed for each speaker.

FIG. 13 is a block diagram of the font converter according to the present embodiment. The same blocks as those of the font conversion apparatus according to the second embodiment are denoted by the same reference numerals, and detailed description will be omitted. The font conversion apparatus according to the present embodiment includes a voice input unit 101, a display unit 103, a voice recognition unit 10
5. Font designation input section 107, speaker identification label acquisition section 201, speaker characteristic learning section 202, speaker identification section 203,
Font conversion control unit 204, font table creation unit 2
05, a font table reference unit 206, a speaker identification feature extraction unit 208, and a font table 1002.

The speaker identification label acquiring section 201 is composed of an input device such as a keyboard, and inputs a speaker identification label such as a speaker name. Speaker identification feature extraction unit 208
When audio data is input from the audio input unit 101,
This is to extract a voice feature amount such as vocal tract information that well captures the characteristics of a speaker from input voice data.

The speaker feature learning unit 202 receives a speech feature amount for each speaker from the speaker identification feature extraction unit 208 and a speaker identification label from the speaker identification label acquisition unit 201. , Using the speaker identification label and the speech feature amount,
A template used for speaker identification is created, and the template is output to the speaker identification unit 203.

The speaker identification unit 203 uses the speech feature amount for speaker identification input from the speaker identification feature extraction unit 208 and the template created by the speaker feature learning unit 202 to identify the speaker. The identification is performed, and a speaker identification label as a recognition result is provided to the font conversion control unit 204.

The font table creation unit 205 newly creates the font table 1002 using the speaker identification label input from the speaker identification label acquisition unit 201 and the font information input from the font designation input unit 107. To create. FIG. 14 shows an example of the font table 1002. The font table as shown in FIG. 14 may be completed by inputting the speaker identification label into the font table in which the font information has already been input.

When the speaker identification label is input from the font conversion control unit 204, the font table reference unit 206 refers to the font table 1002 and outputs font information corresponding to the input speaker identification label. Things. The font conversion control unit 204 controls the voice recognition unit 10
When text data is input from the speaker 5, the speaker identification unit 20
3 is given to the font table reference unit 206, and based on the font information input from the font table reference unit 206, the speech recognition unit 10
The font data is converted from the text data input from step 5. The display unit 103 displays text (sentence) of a standard font before font conversion, and displays text whose font has been converted by the font conversion control unit 204.

The operation of the thus configured font conversion apparatus according to the present embodiment will be described. In the present embodiment, a template used for speaker identification is created before actual text is input by voice, and the font table 10
02 is also created. Then, when the speaker performs voice input, the speaker identification unit 203 identifies the input speaker. The voice recognition unit 105 performs voice recognition on the input voice data and outputs the data to the font conversion control unit 204. The font conversion control unit 204 converts the text data of the speech recognition result into fonts based on font information defined for each speaker.

The above operation example will be specifically described. First,
Before the speaker inputs the actual text by voice, the speaker identification label 203 needs to be input by the speaker identification label acquisition unit 201 in order to create a template used by the speaker identification unit 203. Therefore, as shown in FIG. 14, the registration label is entered with the name of the speaker, such as ICHIRO, JIRO... GORO. Further, the speaker utters a test text, and a speaker identification feature extracting unit 20
8 is input with audio data. As a specific example, consider a case in which two speakers, Ichiro and Jiro, input their test texts by voice. First, Ichiro inputs a label “ICHIRO” from the speaker identification label acquiring unit 201 to the speaker characteristic learning unit 202. After this, Ichiro utters "Ichiro" and reads out other test sentences at normal sound pressure. Speech data “Ichiro” and other test speech data are provided from the speech input unit 101 to the speaker identification feature extraction unit 208.

The speaker identification feature extraction unit 208 extracts the speech feature amount used for speaker identification using the speech data “Ichiro” and other speech data input from the speech input unit 101, and extracts this speech feature. The amount is output to the speaker characteristic learning unit 202. The speaker feature learning unit 202 includes a speaker identification feature extraction unit 2
08, a template for the label “ICHIRO” input from the speaker identification label acquiring unit 201 is created. In this way, an ICHIRO template is created, and a JIRO template is created in the same manner.

Next, a font table 1002 is created. Before inputting the actual text by voice, the speaker inputs the speaker identification label to the font table creation unit 205 via the speaker identification label acquisition unit 201. In the above example, “ICHIRO”,
The label “JIRO” is displayed in the font table creation unit 205
, The state of the font table 1002 is as shown in FIG.

Further, the speaker inputs a font used by the speaker from the font designation input unit 107 to the font table creation unit 205. Font table creation unit 205
The font table 1002 uses the speaker identification label input from the speaker identification label acquisition unit 201 and the font information input from the font designation input unit 107.
Create In the above example, the font designation input unit 107 for the label “ICHIRO” of the speaker in FIG.
To the font table creation unit 205 from “(Font) Gothic, (Style) bold, (Size) 32,
No (text decoration), (color) black "is input. Also, for the speaker label "JIRO", "(Font) Gothic,
(Style) bold italic, (size) 24, (text decoration)
None, (color) black "is input. Thus, FIG.
Is created.

Next, the speaker speaks the actual text. In FIG. 13, voice data is input from a voice input unit 101 to a speaker identification feature extraction unit 208 and a voice recognition unit 105. In the above example, suppose that Ichiro uttered, "Tomorrow will be fine." Then, Ichiro's voice data “Tomorrow will be fine.” Is output via the voice input unit 101 to the speaker identification feature extraction unit 208 and the voice recognition unit 105.
Entered as

The speaker identification feature extraction unit 208 extracts Ichiro's voice feature amount from the speaker identification voice data "I'll be fine tomorrow." Speaker identification unit 2
03, the speech feature amount input from the speaker identification feature extraction unit 208 is compared with the ICHIRO created by the speaker feature learning unit 202.
Compare and collate using JIRO templates to identify speakers. Then, the speaker label “ICHIRO” as the identification result is output to the font conversion control unit 204.

Next, the speech recognition unit 105
After the voice data input from 1 is recognized as "Tomorrow will be fine.", The text data of the recognition result is output to the font conversion control unit 204. In the font conversion control unit 204, the label “ICHIRO” of the speaker input from the speaker identification unit 203 is stored in the font table reference unit 206.
Is output to

The font table reference unit 206 is provided in FIG.
Referring to the font table 1002 of (b), font information “(Font) Gothic, (Style) bold, (Size) 32, (No character decoration) corresponding to the label“ ICHIRO ”input from the font conversion control unit 204 ,(color)
"Black" is obtained and output to the font conversion control unit 204.

The font conversion control unit 204 converts the font of the text "Tomorrow will be fine" input from the speech recognition unit 105 based on the font information input from the font table reference unit 206. Then, the font-converted text is displayed on the display unit 103 as shown in FIG. If another speaker, Jiro, utters a similar sentence, "The day after tomorrow will be rain."
Are displayed on the display unit 103 as shown in FIG.

The font conversion apparatus according to the present embodiment is further provided with a font conversion control unit, a font change input unit, and a font learning unit according to the third embodiment, so that information in the font table 1002 can be updated. It may be.

Further, a text input unit 112 according to the fourth embodiment is further provided for the font conversion apparatus according to the present embodiment, and a reading designation display unit according to the fourth embodiment is used instead of the voice recognition unit 105 according to the fourth embodiment. Using 113, the font of the input text may be converted for each speaker as in the fourth embodiment.

Further, the font conversion apparatus according to the present embodiment is further provided with the text input section 112 and the position information detecting section 115 according to the fifth embodiment, so that the input text can be changed as in the fifth embodiment. May be converted into fonts for each speaker.

Also, as described in the present embodiment, the method of font-converting a text for each speaker can be used in the font converters of the first, third, fourth and fifth embodiments. As a result, the font conversion apparatuses according to the first, third, fourth, and fifth embodiments can be used by a plurality of persons.

As described above, according to the font conversion apparatus of the present embodiment, it is possible to convert the font of the text input by voice according to the speaker who utters and display it on the screen. For example, when a plurality of persons input each sentence, by specifying a font for each speaker, it is possible to make a third party understand who is the speaker who input the sentence.

(Embodiment 7) Next, a font conversion apparatus according to Embodiment 7 of the present invention will be described with reference to the drawings. The font conversion apparatus according to the present embodiment is characterized in that personal information such as sex, emotion, and age is used as voice information of a speaker when performing font conversion on text.

FIG. 16 is a configuration diagram of the font conversion apparatus according to the present embodiment. Here, the same blocks as those in the first embodiment are denoted by the same reference numerals, and detailed description thereof will be omitted. The font conversion apparatus according to the present embodiment includes a voice input unit 1
01, a display unit 103, a personal information extraction unit 301, a personal information recognition unit 302, a font conversion control unit 303, a font table reference unit 305, and a font table 1003.

[0099] The personal information extracting section 301 is provided with the voice input section 10.
When the speaker's voice data is input from 1, the speaker's personal information including the vocal cords and vocal tract information of the speaker is extracted from the input voice data. Such personal information is provided to the personal information recognition unit 302. When the personal information of the speaker is input, the personal information recognition unit 302 recognizes the personal information,
It generates personal status information such as gender, emotion, and age as a recognition result. These pieces of personal state information are provided to the font conversion control unit 303.

The font table 1003 is a table that classifies an individual's state into a plurality of attributes and stores various font setting values for each attribute. Font table 1003
FIG. 17 shows an example. The font table 1003a in FIG. 17A sets gender as an attribute, and FIG.
Is an example of a font table in which emotion is set as an attribute, and a font table 1003c in FIG. 17C is an example in which age is set as an attribute.
The attribute of the personal status information and the font information in the font table 1003 may be input by a speaker or may be given from the beginning.

When the speaker's personal status information is input from the font conversion control unit 303, the font table reference unit 305 refers to the font table 1003, acquires font information corresponding to the personal status information, and performs font conversion control. This is given to the unit 303. Font conversion control unit 3
03, when the personal status information is input from the personal information recognizing unit 302 and the text data of the voice recognition result is input from the voice recognizing unit 105, the personal status information is given to the font table reference unit 305, and the speaker's personal status is displayed. The font information corresponding to the information is obtained, and the text data input from the speech recognition unit 105 is font-converted based on the font information. The display unit 103 displays a text (text) of a standard font before font conversion, and displays the text whose font has been converted by the font conversion control unit 303.

An example of the operation of the thus configured font conversion apparatus according to the present embodiment will be described. When the speaker utters a voice and captures a voice with the voice input unit 101, voice data is output to the personal information extraction unit 301 and the voice recognition unit 105. As an example, suppose an adult man said, "Tomorrow will be fine." First, male voice data “Tomorrow will be fine” is output from the voice input unit 101 to the personal information extraction unit 301 and the voice recognition unit 105.

[0103] The personal information extraction unit 301
The personal information such as vocal cords and vocal tract information is extracted from the voice data "Tomorrow will be fine." The personal information recognizing unit 302 converts the vocal cords and vocal tract information input from the personal information extracting unit 301 into a frequency spectrum, and determines the sex of the speaker from the formant data. Here, it is determined that the low-order formant frequency is low, and the speaker is recognized as "male". The recognition result “male” is output to the font conversion control unit 303.

On the other hand, the voice recognition unit 105 performs voice recognition on the voice data input from the voice input unit 101 as “Tomorrow will be fine.” The recognition result text is output to the font conversion control unit 303. The font conversion control unit 303 provides the personal status information “male” input from the personal information recognition unit 302 to the font table reference unit 305.

The font table reference unit 305 stores, as font information corresponding to the personal status information “male” input from the font conversion control unit 303, for example, as shown in FIG.
The font information “(Font) Gothic,
(Style) bold, (size) 32, (no character decoration), (color) black ”are obtained and output to the font conversion control unit 303.

The font conversion control unit 303 converts the font information “(font) gothic, (style) bold, (size) 3” input from the font table reference unit 305.
2, the text data "Tomorrow will be fine" input from the voice recognition unit 105 is font-converted based on "2, (text decoration), (color) black". Then, the font-converted text is displayed on the display unit 103 as shown in FIG.

[0107] The woman said, "The day after tomorrow will rain."
When the voice input is performed, the same processing as in the above example is performed, and the font-converted text is displayed on the display unit 103 as shown in FIG.

As a specific example, personal information has been considered as gender, but personal information may be a speaker's emotion. If the personal information is emotion, the font of the character input by voice is converted according to the speaker's emotion when speaking.
For example, by changing the color of the character according to the recognized emotion of the speaker, the speaker who is the writer of the text can transmit the emotion to the reader of the text. The font table 1003b used in this case is as shown in FIG. In the case of fun, the font of the displayed text is, for example, “(Font) Gothic, (Style) bold, (Size) 32, (No character decoration), (Color) black”, and in the sad case, it is displayed. The font of the text is, for example, "(Font) Mincho, (Style)
(Size) 32, no (text decoration), (color) blue ".

[0109] The personal information may be age. In this case, the font of the text input by speech is converted based on the estimated age of the speaker who uttered. The font table 1003c used in this case is shown in FIG.
(C). If the user is over 20 years old, the font of the displayed text is, for example, “(Font) Gothic, (Style) bold, (Size) 32, (Character decoration)
None, (color) black ", and if the user is under 20 years old, the font of the displayed text is, for example," (font) Mincho, (style) italic, (size) 14, (character decoration)
None, (color) blue ".

For example, when the personal status is set to age and the child inputs text by voice, the font table 1003
c is used to convert the input text to a font under 20 years old. Then, the reader can identify that the speaker is a child by looking at the converted text. In any case, it is necessary to set in advance whether the individual's condition is gender, emotion, or age.

The font conversion apparatus according to the present embodiment is further provided with a font table creation section and a font designation input section as in the second embodiment, and a new font table 1003 according to the present embodiment can be created. You may do so.

Further, the font conversion apparatus according to the present embodiment is further provided with a font conversion control unit, a font change input unit, and a font learning unit as in the third embodiment.
The font table 1003 of this embodiment may be updated.

Further, the text conversion unit 112 according to the fourth embodiment is further provided in the font conversion device according to the fourth embodiment, and the reading designation display unit according to the fourth embodiment is used instead of the voice recognition unit 105 according to the fourth embodiment. By using 113, the font of the input text may be converted for each speaker.

Further, the font conversion apparatus according to the present embodiment is further provided with a text input unit 112 and a position information detection unit 115 according to the fifth embodiment. May be font-converted for each speaker.

As described above, according to the font conversion apparatus of the present embodiment, it is possible to convert the font of the character input by voice according to the speaker's emotion at the time of uttering, and display it on the screen. For example, by changing the color of characters according to the perceived speaker's emotion, the writer (speaker)
Can convey emotions to readers (readers).

Further, it becomes possible to convert the font of the text input by voice and display it on the screen, depending on the gender of the speaking speaker. For example, by converting the font of the character according to the recognized gender, the reader (reader) of the text can understand at a glance whether the writer (speaker) of the text is male or female.

Further, it becomes possible to convert the font of the character input by voice according to the age of the speaking speaker and display it on the screen. For example, when a child voice-inputs a text, a reader (reader) can understand that the writer (speaker) of the text is a child by reading the converted text.

The personal information extracting section 301 and the personal information recognizing section 302 have the function of the voice information extracting section for extracting the voice information indicating the voice characteristics of the speaker, but according to the content of the voice information. , The following first to third audio information extracting units.

The first voice information extraction unit includes an emotion identification feature extraction unit that extracts, as voice information, a voice feature used for speaker's emotion recognition from the voice data input by the voice input unit, An emotion identification unit that identifies a speaker's emotion using the speech feature amount extracted from the feature extraction unit is included.

The second voice information extraction unit includes a gender identification feature extraction unit for extracting, as voice information, a voice feature used for gender identification of a speaker from the voice data input by the voice input unit, and a gender identification feature extraction unit. It includes a gender identification unit that identifies the gender of the speaker using the speech feature amount extracted from the feature extraction unit.

The third voice information extraction unit includes an age identification feature extraction unit that extracts, as voice information, a voice feature used to identify the age of the speaker from the voice data input by the voice input unit, An age identification unit that identifies the age of the speaker using the audio feature amount extracted from the feature extraction unit is included.

(Eighth Embodiment) Next, a font conversion apparatus according to an eighth embodiment of the present invention will be described with reference to the drawings. The font conversion apparatus according to the present embodiment is characterized in that a font setting value for converting a text into a font is changed based on certainty information for recognizing voice data.

FIG. 19 is a block diagram of the font converter according to the present embodiment. The same blocks as in the first embodiment are denoted by the same reference numerals, and detailed description is omitted. The font conversion device according to the present embodiment includes a voice input unit 101, a display unit 103, a voice recognition unit 401, a font conversion control unit 4
02, a font table reference unit 403, and a font table 1004.

When voice data is input from the voice input unit 101, the voice recognition unit 401 recognizes the voice data, outputs text data as a recognition result, and outputs the reliability information used in recognizing the voice data. (Likelihood information) to be output to the font conversion control unit 402.

The font table 1004 is a table in which a plurality of likelihood information are classified into classes and various font setting values are stored for each class. An example of the font table 1004 is shown in FIG. Here, a class with a likelihood of 3000 or more and a class with a likelihood of less than 3000 are set. Note that the class of the likelihood information of the text in the font table 1004 and the corresponding font information may be input by the speaker or may be provided from the beginning.

The font table reference unit 403 in FIG.
When the likelihood information is input from the font conversion control unit 402, the font conversion unit 4 refers to the font table 1004 and acquires font information corresponding to the likelihood information.
02. Font conversion control unit 402
When the text recognized by the voice recognition unit 401 and the likelihood information of the text are input, the text data is font-converted using font information corresponding to the text likelihood information. The display unit 103 displays text (sentence) in a standard font, and displays font-converted text.

An example of the operation of the thus configured font conversion apparatus according to the present embodiment will be described. Voice data is input from the voice input unit 101 to the voice recognition unit 401. As a specific example, consider a case where a speaker inputs a voice saying "Tomorrow will be fine." First, when the speaker utters “Tomorrow is”, the voice data of “Tomorrow is” is input from the voice input unit 101 to the voice recognition unit 401.

The voice recognition unit 401 performs voice recognition on the input voice data, and converts the voice data into a text "Tomorrow is." It is also assumed that the likelihood of this text “Tomorrow is” is 4000. This text data "Tomorrow is" and likelihood 400
0 is output to the font conversion control unit 402.

The font conversion control section 402 outputs the likelihood 4000 input from the speech recognition section 401 to the font table reference section 403. Font table reference section 4
03, when the likelihood 4000 is input from the font conversion control unit 402, the font table 1004 shown in FIG.
See Then, the font table reference unit 403
The font conversion control unit 4 acquires the font information “(Font) Gothic, (Style) standard, (Size) 20, (No character decoration), (Color) black” corresponding to the likelihood 4000.
02 is output.

The font conversion control unit 402 converts the font of the text data “Tomorrow is” input from the voice recognition unit 401 based on the font information input from the font table 404. Then, the display unit 103 displays the font-converted text as shown in FIG.

Next, it is assumed that the speaker utters “Will be fine.” And the text “Would be fine.” Is recognized with a likelihood of 2000. Thereafter, the same processing as “Tomorrow is performed” is performed, and the display unit 103 changes the text “Will be fine” to “(Font) Gothic, (Style) bold, (Size) as shown in FIG. ) 32, without (text decoration),
(Color) black ".

The font conversion apparatus according to the present embodiment is further provided with a font table creation section and a font designation input section as in the second embodiment, so that the font table 1003 according to the present embodiment is newly provided. May be created.

The font conversion apparatus according to the present embodiment is further provided with a font conversion control section, a font change input section, and a font learning section as in the third embodiment.
The font table 1003 of this embodiment may be updated.

As described above, according to the font converter of the present embodiment, the font of the text input by voice can be converted based on the certainty of the recognized text and displayed on the screen. For example, if the certainty of the recognized text is equal to or less than the reference value, the degree of certainty of the text can be communicated to the reader (reader) by emphasizing the font of the text.

[0135]

According to the font conversion apparatus of the present invention, the font of the text input by voice can be converted based on the voice information of the speaker and displayed on the screen.

Further, according to the font conversion device of the present invention, it is possible to convert the font of the text input by voice according to the strength of the utterance and display it on the screen.
For example, if the power is equal to or higher than a certain power, the normal font and the emphasized font can be selectively used depending on the strength of the voice by using the emphasized font.

Further, according to the font conversion apparatus of the third aspect, it is possible to convert the font of the text input by the uttering speaker and display it on the screen. For example, when a sentence is input by a plurality of persons, by specifying a font for each speaker and inputting the sentence, it is possible to understand who is the speaker who input the sentence.

Further, according to the font conversion apparatus of the fourth aspect, it is possible to convert the font of the text input by voice according to the speaker's emotion at the time of uttering and display the font on the screen. For example, by changing the color of characters according to the perceived speaker's emotion, the writer (speaker)
Can convey the writer's feelings to the reader (reader).

Further, according to the font conversion apparatus of the present invention, it is possible to convert the font of the text input by voice according to the gender of the speaker who utters and display the font on the screen. For example, by converting the font of the character according to the recognized gender, the reader (reader) of the text can understand at a glance whether the writer (speaker) of the text is male or female.

Further, according to the font converting apparatus of the present invention, it is possible to convert the font of the text input by voice according to the age of the speaker who utters and display it on the screen. For example, if a child uses a child font for text input by a child, the reader (reader) may determine that the text writer (speaker) is a child by reading the font-converted text. it can.

Further, according to the font conversion apparatus of the present invention, by learning the voice power information of the individual, it is possible to absorb the difference between the large and small voices.

Further, according to the font conversion apparatus of the present invention, a learning function of correcting and storing the converted font once is obtained. With the learning function, characters can be converted into fonts according to preference.

According to the ninth aspect, when a part of the input text is specified, the specified text is automatically changed by speaking the specified text. Can be.

According to the font conversion apparatus of the tenth aspect, a function of automatically retrieving the text to be changed by recognizing the input voice and comparing it with the already input text is added. . This allows the speaker to read naturally when changing the font.

Further, according to the font conversion apparatus of the present invention, it is possible to convert the font of the text input by voice based on the certainty of the recognized text and display it on the screen. For example, if the certainty of the recognized text is equal to or less than a certain value, the degree of certainty of the text can be communicated to a reader (reader) by emphasizing the text.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a font conversion device according to a first embodiment of the present invention.

FIG. 2 is an example of a font table used in the font conversion device according to the first embodiment.

FIG. 3 is a display example showing the operation of the font conversion device according to the first embodiment.

FIG. 4 is a configuration diagram of a font conversion device according to a second embodiment of the present invention.

FIG. 5 is an example of a font table used in the font conversion device according to the second embodiment.

FIG. 6 is a configuration diagram of a font conversion device according to a third embodiment of the present invention.

FIG. 7 is a display example showing the operation of the font conversion apparatus according to the third embodiment.

FIG. 8 is an example of a font table used in the font conversion device according to the third embodiment.

FIG. 9 is a configuration diagram of a font conversion device according to a fourth embodiment of the present invention.

FIG. 10 is a display example showing an operation of the font conversion apparatus according to the fourth embodiment.

FIG. 11 is a configuration diagram of a font conversion device according to a fifth embodiment of the present invention.

FIG. 12 is a display example showing an operation of the font conversion apparatus according to the fifth embodiment.

FIG. 13 is a configuration diagram of a font conversion device according to a sixth embodiment of the present invention.

FIG. 14 is an example of a font table used in the font conversion device according to the sixth embodiment.

FIG. 15 is an explanatory diagram illustrating a font table creation operation of the font conversion device according to the sixth embodiment.

FIG. 16 is a configuration diagram of a font conversion device according to a seventh embodiment of the present invention.

FIG. 17 is an example of a font table used in the font conversion device according to the seventh embodiment.

FIG. 18 is a display example showing the operation of the font conversion apparatus according to the seventh embodiment.

FIG. 19 is a configuration diagram of a font conversion device according to an eighth embodiment of the present invention.

FIG. 20 is an example of a font table used in the font conversion device according to the eighth embodiment.

FIG. 21 is a display example showing the operation of the font conversion apparatus according to the eighth embodiment.

[Explanation of symbols]

101 voice input unit 102, 109, 204, 303, 402 font conversion control unit 103 display unit 104, 305, 403 font table reference unit 105, 401 voice recognition unit 106 power information extraction unit 107 font designation input unit 108 font table creation unit Reference Signs List 110 font learning unit 111 font change input unit 112 text input unit 113 reading designation display unit 115 position information detection unit 201 speaker identification label acquisition unit 202 speaker characteristic learning unit 203 speaker identification unit 205 font table creation unit 208 speaker Identification feature extraction unit 301 Personal information extraction unit 302 Personal information recognition unit 1001, 1002, 1003, 1004 Font table

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Satoru Inagaki 1006 Kazuma Kadoma, Osaka Pref. Matsushita Electric Industrial Co., Ltd. (72) Inventor Yoshihiro Kojima 1006 Kadoma Kadoma, Kadoma City Osaka Pref. Matsushita Electric Industrial Co., Ltd. 72) Inventor Kazuhiro Koyama 1006 Kazuma Kadoma, Kadoma City, Osaka Prefecture F-term in Matsushita Electric Industrial Co., Ltd. 5B009 KB00 RB31 5D015 AA03 HH01 KK02

Claims (11)

[Claims] 1. A font conversion apparatus for recognizing a voice uttered by a speaker, converting the voice into text data, and converting a font of the text data based on voice information included in the input voice of the speaker. A voice input unit that inputs voice uttered by a speaker and outputs voice data; a voice recognition unit that recognizes voice data input from the voice input unit and converts the voice data into text data; A speech information extraction unit for extracting speech information indicating the speaker's speech characteristics from speech data input by the user, and discriminating the speech information into a plurality of classes or attributes, and a font of a text corresponding to each class or attribute. A font table that stores information, a font table reference unit that refers to the font table and obtains font information corresponding to the input voice information, When the voice information of the speaker is provided from the voice information extraction unit and the text data of the speaker is provided from the voice recognition unit, the voice information is provided to the font table reference unit, and obtained from the font table. A font conversion device, comprising: a font conversion control unit that converts font data of the text data using font information; and a display unit that displays text converted by the font conversion control unit. 2. The power information extracting unit according to claim 1, wherein the voice information extracting unit is a power information extracting unit that extracts a voice power of a speaker as the voice information from voice data input from the voice input unit. The described font converter. 3. A speaker identification feature extraction unit for extracting, as speech information, a speech feature amount used for speaker identification from speech data input by the speech input unit, A speaker identification label acquisition unit that inputs a speaker identification label corresponding to the speaker; a speech feature amount extracted from the speaker identification feature extraction unit; and a speech input from the speaker identification label acquisition unit. A speaker feature learning unit that creates a template used for speaker identification using the speaker identification label; a speech feature amount extracted from the speaker identification feature extraction unit; and a speaker feature learning unit created by the speaker feature learning unit. A speaker identification unit that identifies a speaker using the template and outputs a speaker identification label of the speaker who uttered the voice; and a speaker identification label obtained from the speaker identification label acquisition unit. Font information specified for each speaker And a font table creation unit that creates a font table using,
2. The font conversion apparatus according to claim 1, further comprising: 4. An emotion identification feature extraction unit for extracting, from the audio data input by the audio input unit, an audio feature used for emotion identification of a speaker as the audio information, the audio information extraction unit includes: 2. The font conversion apparatus according to claim 1, further comprising: an emotion identification unit that identifies a speaker's emotion using the voice feature amount extracted from the emotion identification feature extraction unit. 5. The gender identification feature extraction unit that extracts, as speech information, a speech feature used for gender identification of a speaker from the speech data input by the speech input unit, 2. The font conversion apparatus according to claim 1, further comprising: a gender identification unit that identifies the gender of the speaker using the voice feature extracted from the gender identification feature extraction unit. 6. An age-identifying feature extracting unit that extracts, as the speech information, a speech feature amount used for speaker age discrimination from the speech data input by the speech input unit, 2. The font conversion device according to claim 1, further comprising: an age identification unit that identifies the age of the speaker by using a speech feature amount extracted from the age identification feature extraction unit. 7. A new font table for storing font information corresponding to each class of the voice feature amount is generated, and a voice uttered by a speaker is recognized and converted into text data, and the input voice of the speaker is converted to text data. A font conversion device for converting a font of the text data based on included voice information using the font table, wherein the voice input unit inputs a voice uttered by a speaker and outputs voice data, A voice recognition unit that recognizes voice data input from a voice input unit and converts the voice data into text data; and voice information that extracts voice information indicating voice characteristics of a speaker from the voice data input by the voice input unit. An extraction unit, a font designation input unit for inputting font information, and a phonebook of text corresponding to each class, discriminating the voice information into a plurality of classes. A font table storing information, and using the font information input by the font designation input unit and the audio information extracted by the audio information extraction unit, font information corresponding to each audio information is stored in the font table. A font table creating unit for newly storing the input audio information by referring to the font table; a font table reference unit for obtaining font information corresponding to the input audio information; and audio information of the speaker from the audio information extracting unit. A font conversion unit that, when text data of a speaker is provided from the voice recognition unit, provides the voice information to the font table reference unit, and performs font conversion on the text data using font information obtained from the font table; A control unit, and a display for displaying the text converted by the font conversion control unit Font conversion device characterized by the, the equipped. 8. Recognizing a voice uttered by a speaker, converting the voice into text data, converting a font of the text data based on voice information included in input voice of the speaker, and changing a font of the text. Is executed, the font conversion device uses the changed font for subsequent font conversion, wherein a voice uttered by a speaker is input, and a voice input unit that outputs voice data; and A voice recognition unit that performs voice recognition on input voice data and converts the voice data into text data, and a voice information extraction unit that extracts voice information indicating voice characteristics of a speaker from the voice data input by the voice input unit. A font table for discriminating the audio information into a plurality of classes, and storing text font information corresponding to each class; A font table reference unit for acquiring font information corresponding to input voice information by referring to a text file; a speaker inputting text information including a text to be font-changed, a display position of the text, and font information after the font change. A font change input unit to be provided, when the voice information of the speaker is provided from the voice information extraction unit, and when the text data of the speaker is provided from the voice recognition unit, voice information is provided to the font table reference unit; Using the font information obtained from the font table, the text is font-converted, and voice information used when the text input from the font change input unit is font-converted is output, and the voice information is input from the font change input unit. The changed text to be changed is font-formatted using the font information input from the font change input section. A font conversion control unit for changing fonts, a font learning unit for changing data of the font table based on audio information output from the font conversion control unit and font information input from the font change input unit, A display unit for displaying text converted by the font conversion control unit. 9. A font conversion device for extracting speech information of a speaker for each input text data by speaking the already input text data by the speaker, and converting a font of the input text data. A voice input unit that inputs voice uttered by a speaker and outputs voice data; and a voice that extracts voice information indicating a voice characteristic of the speaker from the voice data input by the voice input unit. An information extraction unit, a font table that discriminates the voice information into a plurality of classes, and stores text font information corresponding to each class; and, by referring to the font table, font information corresponding to the input voice information. A font table reference section to be acquired; a text input section for inputting text information including text and a text display position; A part of the text input by the input unit is specified, and a speaker is instructed to speak the specified text, and a text-to-speech specification display unit that outputs text information of the specified text; and When voice information of a speaker is given and text information is given from the reading designation display unit, the speech information is given to the font table reference unit, and the reading designation is used by using the font information obtained from the font table. A font conversion control unit for performing font conversion on the text specified on the display unit; and a display unit for displaying the text input by the text input unit and displaying the text converted by the font conversion control unit. A font conversion device characterized by the above-mentioned. 10. A font converter for extracting speaker's voice information for each input text data by uttering the already input text data by the speaker, and converting the font of the input text data. A voice input unit that inputs voice uttered by a speaker and outputs voice data; a voice recognition unit that recognizes voice data input from the voice input unit and converts the voice data into text data; A speech information extraction unit for extracting speech information indicating a speaker's speech characteristics from speech data input by the speech input unit; and a speech font for discriminating the speech information into a plurality of classes and corresponding to each class. A font table that stores information; and a font table reference unit that refers to the font table and obtains font information corresponding to input voice information. A text input unit for inputting text information including a text and a text display position; comparing the text recognized by the voice recognition unit with the text input from the text input unit; A position information detection unit to be output; and, when speech information of the speaker is given from the speech information extraction unit, and when text information spoken by the speaker is given from the position information detection unit, the speech is referred to the font table reference unit. Giving information, using the font information obtained from the font table, a font conversion control unit for performing font conversion of the text output from the position information detection unit, and displaying the text input by the text input unit, A display unit for displaying the text converted by the font conversion control unit. Font conversion device that. 11. A font conversion device for recognizing a voice uttered by a speaker and converting it into text data, and converting a font of the text data based on the likelihood of the voice recognition, wherein A voice input unit for inputting voice data and outputting voice data, and converting voice data input from the voice input unit into text data by voice recognition, and likelihood information on the likelihood of the voice-recognized text. And a speech recognition unit that outputs the likelihood information into a plurality of classes, and a font table that stores font information corresponding to each class. A font table reference unit for acquiring font information of the text, and text data of the speaker and text likelihood information from the speech recognition unit. The font conversion control unit that gives the likelihood information to the font table reference unit and converts the text into a font using the font information obtained from the font table. A display unit for displaying a text.