JP5954221B2 - Sound source identification system and sound source identification method - Google Patents

Description

  The present invention relates to a sound source specifying system and a sound source specifying method for specifying sound source data suitable for generating synthesized sound based on text data.

2. Description of the Related Art Conventionally, a speech synthesizer that reads out input text data using a known speech synthesis technique is known (see Patent Document 1).
In the speech synthesizer described in this Patent Document 1, the text represented by the input sentence data is analyzed, and attribute information is derived as the analysis result. Then, the attribute information which is the analysis result is compared with the prosodic parameters previously associated with the attribute information, and the speech synthesis is performed using the prosodic parameters associated with the attribute information whose similarity is equal to or higher than the reference value. .

  The attribute information described in Patent Document 1 is information representing the structure of a sentence, and is information such as the number of mora, accent type, part of speech, etc., for example.

JP 2000-056788 A

By the way, in the speech synthesizer, it is required to give a facial expression suitable for the content of the text to the synthesized speech read out by the speech synthesis.
However, in the speech synthesizer described in Patent Literature 1, prosodic data used for speech synthesis is specified in accordance with attribute information representing a sentence structure. For this reason, in the speech synthesizer described in Patent Document 1, there is a problem that a synthesized sound read out by text synthesis is not given a facial expression suitable for the text.

  In other words, in the conventional technology, when outputting a synthesized sound obtained by reading out text data by speech synthesis, it is difficult to specify sound source data (speech parameters) that can give an appropriate expression to the synthesized sound of the text data. There is a problem that there is.

  Therefore, the present invention has an object to specify sound source data (speech parameters) that can give an appropriate expression to the synthesized sound of the sentence data when outputting the synthesized sound obtained by reading out the sentence data by speech synthesis. .

The sound source identification system of the present invention made to achieve the above object includes a sentence acquisition means, a sentence analysis means, a sound source analysis means, a matching means, and an information presentation means.
In the sound source identification system of the present invention, the sentence acquisition unit acquires sentence data representing a character string constituting the designated sentence, and the sentence analysis unit extracts the sentence represented by the sentence data acquired by the sentence acquisition unit. Analyze and derive a text expression distribution representing the distribution degree of each type of expression that appears in the sentence.

  Then, the sound source analysis means includes at least one voice parameter of the sound uttered for the specified content sentence specified as the text of the content in which a plurality of types of facial expressions appear, and facial expressions representing each expression when the specified content sentence is uttered Each sound source data is acquired and analyzed from a storage device storing sound source data that is data associated with each utterance, and each sound source data is converted into a sound represented by a sound parameter included in the sound source data. A sound source facial expression distribution representing the distribution degree of each type of facial expression to be expressed is derived.

  Further, in the sound source identification system of the present invention, the matching means collates the text expression distribution derived by the sentence analysis means with each of the sound source expression distributions derived by the sound source analysis means and derives a correlation value between them. Then, the information presenting means presents sound source data corresponding to the correlation value having the highest value among the correlation values derived by the matching means.

  According to the sound source identification system of the present invention, it is possible to identify sound source data corresponding to a sound source facial expression distribution having a maximum correlation value with the text facial expression distribution. The sound source data having the maximum correlation value includes a speech parameter having a facial expression distribution with the highest degree of matching in the facial expression distribution appearing in the text data.

  Therefore, according to the sound source identification system of the present invention, when outputting a synthesized sound obtained by reading out text data by speech synthesis, sound source data (speech parameters) that can give an appropriate expression to the synthesized sound of the text data is identified. can do.

The “expression” in the present invention is a concept including at least emotions, emotions, scenes, and situations.
By the way, the sound source identification system of the present invention may include content information acquisition means, waveform acquisition means, parameter derivation means, facial expression data generation means, and sound source data registration means.

  In this case, the content information acquisition means acquires the specified content sentence representing the character string constituting the text of the content in which multiple types of facial expressions appear, and the waveform acquisition means is the specified content sentence acquired by the content information acquisition means. A target waveform, which is a speech waveform uttered for a character string represented by the specific content information, is acquired. Further, the parameter deriving unit derives a speech parameter from the target waveform acquired by the waveform acquiring unit, and the facial expression data generating unit estimates the facial expression expressed in the target waveform based on the specific content information, and the estimation The result is generated as facial expression data.

  Then, the sound source data registration unit generates sound source data by associating the voice parameter derived by the parameter deriving unit with the facial expression data generated by the facial expression data generation unit, and stores the sound source data in the storage device.

According to such a sound source identification system, sound source data can be generated based on utterance content information and a speech waveform corresponding to the utterance content information.
That is, according to the sound source identification system of the present invention, the speech parameters of various speakers are derived by deriving the speech parameters from each target waveform obtained by causing many people to utter the character string represented by the utterance content information. it can.

  As a result, according to the sound source identification system, the types of speech parameters can be diversified, so that when a synthesized sound obtained by reading out text data by speech synthesis is output, it can be selected from various sound source data and a more appropriate expression can be given. .

The present invention may be implemented as a sound source specifying method that is a method of specifying sound source data.
In this case, the sound source localization methods, and text acquisition step of Ru to acquire text data to the computer, to analyze the text represented by the text data, and text analysis process of Ru is derived text expression distribution computer, speech parameters and expressions from a storage device and a data storing sound source data which associates each utterance by acquiring the respective sound source data analyzed, a sound source analysis process of Ru is derived sound source expression distribution computer, the text expression distribution to each sound source expression distribution matching to a matching process which Ru is derived correlation values of both the computer, in the correlation values, have an information presentation step of Ru is presented sound source data corresponding to the highest correlation value is the value to the computer Also good.

  By executing such a sound source specifying method, the same effect as that of the sound source specifying system according to claim 1 can be obtained.

It is a block diagram which shows schematic structure of a sound source identification system. It is a flowchart which shows the process sequence of a sound source data registration process. It is a flowchart which shows the process sequence of a sound source specific process. It is explanatory drawing which shows the process outline | summary of a sound source identification process. It is explanatory drawing which shows the process outline | summary of a sound source identification process.

Embodiments of the present invention will be described below with reference to the drawings.
<Speech synthesis system>
A speech synthesis system 1 shown in FIG. 1 is a system that outputs the contents of text data WT designated by a user as synthesized sounds having characteristics designated by the user, and includes at least one information processing server 10 and at least one speech. And an output terminal 60.

  In this speech synthesis system 1, the information processing server 10 analyzes the text data WT designated by the user of the speech output terminal 60, and at least matches the user's wish from a plurality of pre-registered sound source data SD. The sound source data SD is extracted and presented. Further, in the speech synthesis system 1, the speech output terminal 60 performs speech synthesis based on the sound source data SD, and outputs a synthesized sound having contents corresponding to the designated text data WT.

That is, the speech synthesis system 1 functions as a sound source identification system in the present invention.
<Audio output terminal>
The audio output terminal 60 includes a communication unit 61, an information receiving unit 62, a display unit 63, a sound input unit 64, a sound output unit 65, a storage unit 66, and a control unit 70. As the audio output terminal 60 in the present embodiment, for example, a known portable terminal may be assumed, or a known information processing apparatus such as a so-called personal computer may be assumed. Note that portable terminals include well-known electronic book terminals, and portable information terminals such as mobile phones and tablet terminals.

  In the communication unit 61, the audio output terminal 60 performs information communication with the outside via a communication network. The information receiving unit 62 receives information input via an input device (not shown). The display unit 63 displays an image based on a signal from the control unit 70.

  The sound input unit 64 is a device that converts sound into an electric signal and inputs the electric signal to the control unit 70, and is, for example, a microphone. The sound output unit 65 is a known device that outputs sound, and includes, for example, a PCM sound source and a speaker. The storage unit 66 is a non-volatile storage device configured to be able to read and write stored contents. The storage unit 66 stores various processing programs and various data.

The control unit 70 is configured around a known computer having at least a ROM 72, a RAM 74, and a CPU 76.
That is, each voice output terminal 60 acquires text data WT designated by the user of the voice output terminal 60 and sound source data SD suitable for the text data WT from the information processing server 10 and executes voice synthesis. Then, by the voice synthesis, a synthesized sound representing the contents of the text data WT is generated and output.
<Information processing server>
The information processing server 10 includes a communication unit 12, a control unit 20, and a storage unit 30, and includes at least sentence data WT representing a character string constituting a sentence and at least a voice feature amount of speech input in advance. This is a server in which the sound source data SD is stored.

In the communication unit 12, the information processing server 10 communicates with the outside through a communication network. The communication network in this embodiment is, for example, a public wireless communication network or a network line.
The control unit 20 includes a ROM 22 that stores processing programs and data that need to retain stored contents even when the power is turned off, a RAM 24 that temporarily stores processing programs and data, and processes stored in the ROM 22 and RAM 24. A known computer having at least a CPU 26 that executes various processes according to a program is mainly configured. The control unit 20 controls the communication unit 12 and the storage unit 30.

  The storage unit 30 is a non-volatile storage device configured to be able to read and write stored contents. The storage device is, for example, a hard disk device or a flash memory. The storage unit 30 stores text data WT and sound source data SD.

The text data WT here is, for example, data obtained by converting a book into text data, and is prepared in advance for each book. Books here are novels and the like.
The sound source data SD is data in which the sound parameter PV _j and the tag data (expression data) TG _j are associated with each sound source j.

  The voice parameter PV is at least one feature amount representing a waveform of a sound emitted by a person. This feature amount is a feature amount of speech used for so-called formant synthesis, and is prepared for each speaker and for each phoneme. As a feature quantity in the speech parameter PV, at least a fundamental frequency F0, a mel frequency cepstrum (MFCC), a phoneme length, a power, and a time difference thereof in each phoneme in the uttered speech are provided.

  The tag data TG is data representing the nature of the sound represented by the speech parameter PV, and includes speaker feature data representing the features of the speaker, and expression data representing the expression of the speaker when the speech is spoken. At least. The speaker characteristic data includes, for example, the sex and age of the speaker. The expression data is data representing a concept as an expression including at least emotions, emotions, scenes, and situations, and may include information necessary for estimating the expression of the speaker.

The sound source data SD in which the sound parameters PV and the tag data TG are associated is generated by the sound source data registration process performed by the control unit 20 and stored in the storage unit 30. <Sound source data registration process>
When the sound source data registration process is started, as shown in FIG. 2, sentence data WT representing a character string of a prescribed content sentence prescribed in advance as a sentence having contents in which a plurality of types of facial expressions appear is acquired (S110). ). The text data WT acquired in S110 is hereinafter referred to as utterance content text data.

  Subsequently, one speech waveform data corresponding to the utterance content sentence data obtained in S110 is obtained (S120). The speech waveform data is data representing each speech waveform uttered in advance with respect to the specified content sentence represented by the utterance content text data, and is uttered in advance by various persons.

  Further, a speech parameter SV is derived from each speech waveform data acquired in S120 (S130). In S130 of the present embodiment, the fundamental frequency, the mel frequency cepstrum (MFCC), the power, and the time difference between them are each derived as a voice parameter SV. Since these fundamental frequency, MFCC, and power derivation methods are well known, detailed description thereof is omitted here. For example, if the fundamental frequency is used, autocorrelation along the time axis, autocorrelation of the frequency spectrum, Alternatively, it may be derived using a method such as a cepstrum method. In the case of the MFCC, the result of frequency analysis (for example, FFT) for each time analysis window may be derived by further frequency analysis of the logarithmic magnitude of each frequency. The power may be derived by integrating the squared result of the amplitude in the time analysis window in the time direction.

  Subsequently, in the sound source data registration processing, facial expression data estimation processing for estimating facial expression data is executed (S140). In this facial expression data estimation process, the facial expression expressed by the speech waveform data is estimated based on the result of analyzing the utterance content text data acquired in S110.

  The analysis of “speech content text data” here refers to, for example, acquiring word expression information corresponding to each word for each word specified by morphological analysis of text corresponding to the utterance content text data. The word expression information here is information that associates each word with the contents of the expression represented by each word in advance, and is stored in advance in the word expression database. Then, according to the acquired word facial expression information, the appearance frequency of facial expressions representing the same content is aggregated for each facial expression content, and as a result of the aggregation, the content of the most frequent facial expression is expressed by the facial expression represented by the speech waveform data. Can be estimated.

  Subsequently, by associating the voice parameter SV derived in S130 with the facial expression data estimated in S140 for each corresponding voice waveform data, voice parameter registration for generating the sound source data SD and storing it in the storage unit 30 is performed. Execute (S150). Note that the data associated with the speech parameter SV stored in the storage unit 30 in S150 of the present embodiment includes the content (type) of the spoken sentence, the speaker ID, and the speaker feature data in addition to the facial expression data. (That is, tag data TG). For the speaker ID and speaker feature data, information used for logging in to the information processing server 10, the voice output terminal 60, and other terminals may be acquired as the speaker ID and speaker feature data.

Thereafter, the voice parameter registration process is terminated.
That is, in the speech parameter registration process of the present embodiment, one speech waveform data uttered with respect to the sentence represented by the utterance content sentence data is analyzed, and the speech parameter SV is derived. At the same time, in the voice parameter registration process, the sentence represented by the utterance content sentence data is analyzed, and facial expression data representing the expression expressed by the voice parameter SV is derived.

In the sound parameter registration process, the sound source data SD is generated by associating the corresponding sound parameters SV with the expression data, and the sound source data SD is stored in the storage unit 30. Thus, the storage unit 30 stores the sound source data SD created for each voice uttered with respect to the specified content sentence.
<Sound source identification processing>
Next, the sound source identification process executed by the control unit 20 of the information processing server 10 will be described.

  When this sound source identification process is started, as shown in FIG. 3, the sentence designation information representing the sentence data WT designated by the voice output terminal 60 is acquired (S310). Subsequently, sentence data (hereinafter referred to as “designated sentence data”) WT corresponding to the sentence designation information obtained in S310 is obtained from the storage unit 30 (S320). The designated sentence data WT acquired in S320 is a character string itself constituting the sentence, that is, text data, as shown in FIG.

  Further, the designated sentence data WT acquired in S320 is subjected to text analysis, a character i appearing in the sentence represented by the designated sentence data WT, and a utterance text representing the content of the text that each character i should utter, Is generated in correspondence with the speaker text (S330). In addition, the character i here includes a speaker and a narrator. For example, for a conversational sentence, a speaker representing the person who uttered the conversational sentence in the sentence is identified as the character i, and for a local sentence, the narrator is identified as the character i.

  Specifically, in S330, first, the designated sentence data WT acquired in S310 is divided by punctuation marks and parentheses in the sentence represented by the designated sentence data WT, as shown in FIG. Then, it is divided into utterance texts which are unit sections constituting the sentence. Then, morphological analysis and dependency analysis are performed on the uttered text, and the character i who should utter the unit section is specified. Further, by associating each utterance text with the character i corresponding to the utterance text, the character i (in the figure: speaker) and the utterance text (in the figure: text) as shown in FIG. ) Is generated in correspondence with speaker text.

  For morphological analysis and dependency analysis, a known method may be used. For example, in the case of morphological analysis, “MeCab” may be used. For dependency analysis, “Cabocha (“ Taku Kudo, Yuji Matsumoto, “Japanese Dependency Analysis by Chunking Stage Application” ”, Transactions of Information Processing Society of Japan, 43 (6), 1834-1842 (2001). ")" Or the like may be used.

  Returning to the sound source identification process, the utterance text associated with each character i is analyzed based on the speaker text correspondence data, and the facial expression appearing in each utterance text is identified (S340). The analysis in S340 may be performed by acquiring the content of the facial expression represented by each word included in the utterance text based on the word facial expression information described above.

  Subsequently, for each character i in the text represented by the designated text data WT, the results of facial expression analysis in S340 are tabulated, and a text expression distribution tpd (i, k) representing the distribution of facial expressions for each character i is obtained. Derived (S350). The text expression distribution tpd (i, k) derived in S350 includes items representing each expression to be expressed by the character i in the sentence represented by the designated sentence data WT, as shown in FIG. k represents the distribution of the intensity of each facial expression.

  Further, based on the sound source data SD stored in the storage unit 30, a sound source expression distribution vpd (j, k) representing the expression distribution in each sound source data SD is derived (S360). In S360, specifically, the voice parameter sp_n (j) associated with the facial expression data in which the content of the facial expression is in a neutral state is used as a reference, and the expression data corresponding to each facial expression k is determined from that reference. A vector to each of the attached speech parameters sp_e (j) is derived as a sound source expression distribution vpd (j, k) for each sound source j by the following equation (1).

この（１）式によって音源ｊごとに導出される音源表情分布ｖｐｄは、図５（Ｂ）に示すように、規定内容文にて出現する各表情を項目ｋとして、各表情の強さの分布を表したものとなる。

As shown in FIG. 5B, the sound source expression distribution vpd derived for each sound source j by the expression (1) is a distribution of the intensity of each expression with each expression appearing in the specified content sentence as an item k. It is a representation.

  Subsequently, the text expression distribution tpd (i, k) for each character i derived in S350 is collated with each of the sound source expression distributions vpd (j, k) derived in S360, and the correlation value cor (i, j ) Is derived (S370). The derivation of the correlation value cor (i, j) in S370 is executed according to the following equations (2) and (3).

  Note that TP and VP in the expressions (2) and (3) are the arithmetic mean of the text expression distribution tpd and the sound source expression distribution vpd with respect to the item k of each appearing expression, and kmax is the expression. Is the quantity.

さらに、Ｓ３７０にて導出した相関値ｃｏｒ（ｉ，ｊ）が最大となる音源データＳＤを、登場人物ｉごとに提示する（Ｓ３８０）。このＳ３８０における提示とは、通信部１２を介して、相関値ｃｏｒ（ｉ，ｊ）が最大となる登場人物ｉごとの音源データＳＤを、音声出力端末６０の表示部６３に出力することでも良い。

Furthermore, the sound source data SD that maximizes the correlation value cor (i, j) derived in S370 is presented for each character i (S380). The presentation in S380 may be output to the display unit 63 of the audio output terminal 60 via the communication unit 12 for the sound source data SD for each character i having the maximum correlation value cor (i, j). .

Thereafter, the sound source identification process is terminated.
In other words, in the sound source identification process, the sentence represented by the designated sentence data WT is analyzed, and the text expression distribution tpd (i, k) representing the distribution degree of each type of expression appearing in the sentence for each character i. To derive. Each of the sound source data SD stored in the storage unit 30 is acquired and analyzed, and for each sound source data SD, each type of facial expression expressed in the sound represented by the sound parameter PV included in the sound source data SD. The sound source facial expression distribution vpd (j, k) representing the distribution degree of is derived.

Further, in the sound source identification processing, the derived text expression distribution tpd (i, k) for each character i is collated with each of the sound source expression distributions vpd (j, k), and the correlation value cor (i, j) between the two. ) And the sound source data SD having the highest correlation value cor (i, j) is presented.
[Effect of the embodiment]
As described above, according to the speech synthesis system 1, the sound source corresponding to the sound source expression distribution vpd (j, k) having the maximum correlation value cor (i, j) with the text expression distribution tpd (i, k). Data SD can be specified and presented. The sound source data SD having the maximum correlation value cor (i, j) includes the speech parameter PV having the facial expression distribution with the highest matching degree in the facial expression distribution appearing in the text data WT.

  Therefore, according to the speech synthesis system 1, when outputting a synthesized sound obtained by reading out the text data WT by speech synthesis, the sound source data SD that can give an appropriate expression to the synthesized sound of the text data WT can be specified. it can.

  Moreover, according to the speech synthesis system 1, the sound source data SD can be generated based on the utterance content sentence data and the speech waveform for the utterance content sentence data. The speech synthesis system 1 derives the speech parameters PV from various target waveforms obtained by causing many people to utter the character string represented by the utterance content text data, thereby obtaining the speech parameters PV of various speakers. Can be derived.

As a result, according to the speech synthesis system 1, since the types of speech parameters PV can be diversified, when outputting synthesized speech obtained by reading out the text data WT by speech synthesis, it can be selected from various sound source data SD, and more appropriate Can add facial expressions.
[Other Embodiments]
As mentioned above, although embodiment of this invention was described, this invention is not limited to the said embodiment, In the range which does not deviate from the summary of this invention, it is possible to implement in various aspects.

  That is, the aspect which abbreviate | omitted a part of structure of the said embodiment as long as the subject could be solved is also embodiment of this invention. Further, an aspect configured by appropriately combining the above embodiment and the modification is also an embodiment of the present invention. Moreover, all the aspects which can be considered in the limit which does not deviate from the essence of the invention specified by the wording described in the claims are the embodiments of the present invention.

  For example, in S370 in the sound source identification process of the above embodiment, the sound source data SD having the maximum correlation value cor (i, j) is presented. However, the sound source data SD presented in S370 is not limited to this. Absent. In other words, the sound source data SD presented in S370 may be the sound source data SD corresponding to the character i having a correlation value cor (i, j) from the maximum value to a specified number (eg, “5”) for each character i. good.

  In addition, in the presentation of the sound source data SD in S370, the output target may be determined according to the gender of the character i. That is, if the gender of the character i is male, the sound source data SD in which the gender in the speaker characteristic data is male is presented. If the gender of the character i is female, the gender in the speaker characteristic data is displayed. It is preferable to present the sound source data SD that is female.

In this case, the gender of the character i may be determined from the gender in the proper noun representing the name of the character i or may be determined from the pronoun representing the character i.
In S360 in the sound source identification process in the above embodiment, the text expression distribution tpd (i, k) and the sound source expression distribution vpd (j, k) are normalized when the correlation value cor is derived. Normalization may be performed when each of the text expression distribution tpd (i, k) and the sound source expression distribution vpd (j, k) is derived.
[Correspondence between Embodiment and Claims]
Finally, the relationship between the description of the above embodiment and the description of the scope of claims will be described.

  The function obtained by executing S310 and S320 in the sound source identification processing of the above embodiment corresponds to the sentence acquisition means in the claims, and S330 to S350 are sentence analysis means in the claims. It corresponds to. Further, S360 in the sound source identification process of the above embodiment corresponds to the sound source analysis means in the claims, S370 corresponds to the matching means, and S380 corresponds to the information presentation means.

  In addition, the function obtained by executing S110 in the sound source registration process of the above embodiment corresponds to the content information acquisition unit in the claims, and S120 corresponds to the waveform acquisition unit. The function obtained by executing S130 corresponds to the parameter deriving unit in the claims, S140 corresponds to the facial expression data generation unit, and S150 corresponds to the sound source data registration unit.

  DESCRIPTION OF SYMBOLS 1 ... Speech synthesis system 10 ... Information processing server 12 ... Communication part 20 ... Control part 22 ... ROM 24 ... RAM 26 ... CPU 30 ... Memory | storage part 60 ... Voice output terminal 61 ... Communication part 62 ... Information reception part 63 ... Display part 64 ... Sound input unit 65 ... Sound output unit 66 ... Storage unit 70 ... Control unit 72 ... ROM 74 ... RAM 76 ... CPU

Claims (3)

Sentence acquisition means for acquiring sentence data representing a character string constituting the specified sentence;
Sentence analysis means for analyzing a sentence represented by sentence data acquired by the sentence acquisition means and deriving a text expression distribution representing a distribution degree of each type of expression appearing in the sentence;
For each utterance, at least one voice parameter of a sound uttered for a prescribed content sentence defined as a sentence of content in which multiple types of facial expressions appear, and facial expression data representing each facial expression when uttered for the prescribed content sentence Each of the sound source data is acquired and analyzed from a storage device that stores sound source data that is associated data, and each sound source data is represented in the sound represented by the sound parameter included in the sound source data. Sound source analysis means for deriving a sound source facial expression distribution representing the distribution degree of each type of facial expression;
Matching means for deriving a correlation value between the text expression distribution derived by the sentence analysis means and each sound source expression distribution derived by the sound source analysis means,
A sound source specifying system comprising: information presenting means for presenting the sound source data corresponding to the correlation value having the highest value among the correlation values derived by the matching means. Content information acquisition means for acquiring a specified content sentence representing a character string constituting a sentence of content in which multiple types of facial expressions appear;
A waveform acquisition unit that acquires a target waveform that is a speech waveform uttered for a character string represented by the specific content information that is the specified content sentence acquired by the content information acquisition unit;
Parameter deriving means for deriving the speech parameter from the target waveform obtained by the waveform obtaining means;
Facial expression data generating means for estimating the facial expression expressed in the target waveform based on the specific content information and generating the estimation result as the facial expression data;
Sound source data registration means for generating the sound source data by associating the voice parameter derived by the parameter derivation means with the expression data generated by the expression data generation means and storing the sound source data in the storage device. The sound source specifying system according to claim 1. A sentence acquisition process for causing a computer to acquire sentence data representing a character string constituting the specified sentence;
Analyzing a sentence represented by sentence data acquired in the sentence acquisition process, and causing a computer to derive a text expression distribution representing a distribution degree of each type of expression appearing in the sentence; and
For each utterance, at least one voice parameter of a sound uttered for a prescribed content sentence defined as a sentence of content in which multiple types of facial expressions appear, and facial expression data representing each facial expression of the voice uttered for the prescribed content sentence Each of the sound source data is acquired and analyzed from a storage device that stores sound source data that is associated data, and each sound source data is represented in the sound represented by the sound parameter included in the sound source data. A sound source analysis process that causes the computer to derive a sound source facial expression distribution representing the distribution of each type of facial expression;
Matching the text expression distribution derived in the sentence analysis process with the sound source expression distribution derived in the sound source analysis process, and causing the computer to derive a correlation value between the two,
A sound source identification method comprising: an information presentation process for causing the computer to present the sound source data corresponding to the correlation value having the highest value among the correlation values derived in the matching process.

Images