Information processing device, information processing method, and program

Abstract

[Problem] To provide an information processing device, an information processing method, and a program with which it is possible to output, in accordance with a target of a value generated using content, information relating to correction of a configuration element for generating the content. [Solution] Provided is an information processing device comprising a control unit for carrying out: a process for estimating, on the basis of information pertaining to one or more configuration elements configured in order to generate content, a value generated using the content; a process for comparing the estimated estimation value with a target value; and a process for outputting, on the basis of the result of comparison, correction information relating to correction of the configuration element.

Description

[Technical Field]

[0001] The present disclosure relates to an information processing device, an information processing method, and a program. [Background technology]

[0002] In recent years, proposals have been made for cloud services that use AI (artificial intelligence) to automatically generate text based on several input keywords, and software that supports the creation of scenarios by writing the order of scene changes and lines in a story. For example, Patent Document 1 listed below discloses a technology that analyzes the narrative content of stories in various formats such as books and movies, and graphically represents the relationships between a story that interests a user and many other stories, thereby enabling the user to quickly search for and understand similar stories. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Special Publication No. 2014-507699 Summary of the Invention [Problem to be solved by the invention]

[0004] However, when creating a story script (scenario) or when considering the story's setting (characters, locations, etc.), it required the creator's many years of experience to predict the specific values ​​that would be generated by the resulting content (e.g., video).

[0005] Therefore, the present disclosure proposes an information processing device, an information processing method, and a program capable of outputting information relating to the correction of setting elements for content generation according to a target value generated by the content. [Means for solving the problem]

[0006] According to the present disclosure, an information processing device is proposed that includes a control unit that performs the following processes: estimating a value generated by content based on information on one or more setting elements that are set to generate the content; comparing the estimated value with a target value; and outputting correction information regarding correction of the setting elements based on the result of the comparison.

[0007] According to the present disclosure, an information processing method is proposed, which includes a processor estimating a value to be generated by content based on information of one or more setting elements that are set to generate the content, comparing the estimated value with a target value, and outputting correction information regarding correction of the setting elements based on the result of the comparison.

[0008] According to the present disclosure, a program is proposed that causes a computer to function as a control unit that performs the following processes: estimating a value generated by content based on information on one or more setting elements that are set to generate the content; comparing the estimated value with a target value; and outputting correction information regarding correction of the setting elements based on the results of the comparison. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a block diagram illustrating an example of a basic configuration of an information processing device according to an embodiment of the present disclosure. [Figure 2] FIG. 2 is a block diagram showing an example of the configuration of a control unit according to the first embodiment. [Figure 3] 10 is a flowchart showing an example of the flow of an operation process according to the first embodiment. [Figure 4] FIG. 10 is a diagram illustrating learning of the time length for each setting element of a past work according to the first embodiment. [Figure 5] FIG. 10 is a diagram illustrating estimation of the time length for each setting element of the current work according to the first embodiment. [Figure 6] FIG. 10 is a diagram illustrating a case where a setting element to be added is selected from past works according to the first embodiment. [Figure 7] FIG. 10 is a diagram showing an example of a display of a change proposal for a setting element according to the first embodiment. [Figure 8] FIG. 3 is a diagram illustrating a tagging process according to the first embodiment. [Figure 9] FIG. 10 is a diagram showing an example of a screen displaying the extent to which the deletion of a setting element affects a scenario according to the first embodiment. [Figure 10] FIG. 10 is a diagram showing an example of a screen displaying a history of changes made in response to suggestions according to the first embodiment. [Figure 11] FIG. 2 is a diagram illustrating an example of a screen transition in scenario creation support according to the first embodiment. [Figure 12] FIG. 10 is a diagram showing an example of a title and summary determination screen according to the first embodiment. [Figure 13] FIG. 3 is a diagram showing an example of a genre selection screen according to the first embodiment. [Figure 14] FIG. 10 is a diagram showing an example of a character setting screen according to the first embodiment. [Figure 15] FIG. 4 is a diagram showing an example of a location setting screen according to the first embodiment. [Figure 16] FIG. 10 is a diagram showing an example of a setting screen for details (stage props, props, etc.) according to the first embodiment. [Figure 17] FIG. 10 is a diagram showing an example of an editing screen for a beat sheet according to the first embodiment. [Figure 18] FIG. 3 is a diagram showing an example of an edit screen for a developed plot according to the first embodiment. [Figure 19] FIG. 3 is a diagram showing an example of a scenario editing screen according to the first embodiment. [Figure 20] FIG. 10 is a diagram illustrating a case where a person correlation is proposed in units of a correlation diagram as a reference according to the first embodiment. [Figure 21] FIG. 10 is a block diagram showing an example of the configuration of a control unit according to a second embodiment. [Figure 22] 10 is a flowchart showing an example of the overall flow of operation processing according to the second embodiment. [Figure 23]10 is a flowchart showing an example of the flow of an extraction process according to the second embodiment. [Figure 24] 10 is a flowchart showing an example of the flow of an estimation process according to the second embodiment. [Figure 25] FIG. 10 is a diagram showing an example of a character table according to the second embodiment. [Figure 26] FIG. 10 is a diagram illustrating an example of a location table according to the second embodiment. [Figure 27] FIG. 10 is a diagram showing an example of a props / props table according to the second embodiment. [Figure 28] FIG. 10 is a diagram illustrating identical element determination according to the second embodiment. [Figure 29] FIG. 11 is a diagram showing an example of the data configuration of a past work component DB according to the second embodiment. [Figure 30] FIG. 10 is a diagram illustrating an example of attribute definitions of time-series metadata (component elements) according to the second embodiment. [Figure 31] FIG. 10 is a diagram showing an example of a detailed definition of a "sentence element" included in the time-series metadata according to the second embodiment. [Figure 32] 10 is a flowchart showing an example of the flow of a process of determining importance and assigning a label according to the second embodiment. [Figure 33] FIG. 10 is a diagram showing an example of an editing screen on which the importance and labels of components in a scene can be edited according to the second embodiment. [Figure 34] FIG. 11 is a diagram showing an example of a confirmation screen for the importance of each appearance scene of a component according to the second embodiment. [Figure 35] FIG. 10 is a diagram showing an example of an editing screen for a component (stage props) according to the second embodiment. [Figure 36] FIG. 10 is a diagram showing an example of an editing screen for a component (stage props) according to the second embodiment. [Figure 37] FIG. 10 is a diagram showing an example of an editing screen for a component (character) according to the second embodiment. [Figure 38] 10 is a flowchart showing an example of the flow of a location-based search process for past works according to an application example of the second embodiment. [Figure 39] 10 is a flowchart showing an example of the flow of a performer search process according to an application example of the second embodiment. [Figure 40] 10 is a flowchart showing an example of the flow of a process for trial calculation of CG production costs according to an application example of the second embodiment. [Figure 41] 10 is a flowchart showing an example of the flow of a process for trial calculation of shooting costs according to an application example of the second embodiment. [Figure 42] FIG. 10 is a diagram showing an example of a location table of work A according to an application example of the second embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0010] Preferred embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. In this specification and drawings, components having substantially the same functional configurations are designated by the same reference numerals, and redundant description will be omitted.

[0011] The explanation will be given in the following order: 1. Overview 2.Basic configuration 3. First embodiment 3-1.Configuration example 3-2. Operation processing 3-3. Display screen example 3-4.Application Examples 4. Second embodiment 4-1.Configuration example 4-2. Operation processing 4-3. Display screen example 4-4.Application Examples 5. Supplementary Information

[0012] <<1. Overview>> One embodiment of the present disclosure is characterized in that information regarding correction of setting elements for content generation is output in accordance with a predetermined value target generated by the content. In this specification, content includes, for example, videos such as movies, commercials (CMs), dramas, documentaries, animation, and streaming videos, as well as music, plays, and speeches. All of these are assumed to be generated based on a scenario (script). A scenario is text that describes the story of the content. For example, a scenario used for video generation may include "headlines" that describe the location and time of day, "stage directions" that describe the actions of actors and changes in the stage (scene), and "dialogue" that describe the words spoken by actors. Filming is performed according to the description of the scenario, and the scenario is visualized, i.e., a video is generated.

[0013] "Setting elements for content generation" refers to information that forms the basis of a story, such as the setting (time period, location), characters (relationships), and props (stages, props, etc.).

[0014] The "predetermined value generated by the content" may be various, such as the duration of a video that visualizes a scenario, the cost of creating the video (the time and cost of filming, hereinafter referred to as filming cost), or the income from the video (such as box office receipts for a movie, revenue from video playback, etc.).

[0015] (Identifying issues) In the production of movies and anime, a scenario is generally created first, and then filming and animation are produced based on that scenario, but the length of the video is often predetermined. The experience of the scenario writer determines the optimal amount of content to include in the scenario to fit the set length. Also, the budget for filming costs is often set in advance, but whether or not a scenario can be created that fits within the set filming costs also depends on the experience of the producer. In addition, the producer's experience also determines what elements should be incorporated into the scenario to make it popular and profitable.

[0016] If appropriate decisions are not made during the creation of a scenario, major rework, such as changing the content after filming, will be required. Therefore, it is desirable to make more accurate estimates before the video is created, i.e., at the scenario creation stage. Furthermore, with the recent proliferation of video distribution services, even amateur creators with no video production experience can easily create and publish a video from a scenario. Even in such an environment, there may be a target value, such as wanting to create a video of a certain length that is likely to be popular, such as 10 minutes or less. Even inexperienced users can create better content if they can more accurately estimate certain values, such as the duration of the video when it is created, at the scenario creation stage.

[0017] Therefore, this embodiment is characterized by outputting information regarding correction of setting elements for content generation according to the target of a predetermined value resulting from the content. More specifically, this embodiment estimates predetermined values ​​(e.g., duration, shooting costs, or revenue) resulting from the resulting content (e.g., video) at the stage of creating a scenario or determining the story settings, compares the estimated values ​​with target values, and presents the user with corrections to the setting elements (addition or deletion of setting elements) based on the comparison results. This can support the creation of better content.

[0018] <<2.Basic configuration>> Next, a basic configuration of an information processing device 1 that supports content creation according to this embodiment will be described with reference to Fig. 1. Fig. 1 is a block diagram showing an example of the basic configuration of an information processing device 1 according to an embodiment of the present disclosure.

[0019] As shown in FIG. 1, the information processing device 1 includes an input unit 11, a control unit 12, an output unit 13, and a storage unit .

[0020] (Input section 11) The input unit 11 has a function of accepting information input to the information processing device 1. The input unit 11 may be a communication unit that receives information from an external device or an operation input unit that receives operation input by a user. The operation input unit may be realized, for example, by a mouse, a keyboard, a touch panel, a switch, or a microphone (voice input). The input unit 11 according to this embodiment accepts, for example, information (text data) of a scenario in production, input of element information, and target values ​​(e.g., the duration of content, filming costs, profits, etc.). A scenario in production may be, for example, a scenario in which the first act has been written, whereas a story generally consists of three acts (Act 1 - beginning (setting the situation), Act 2 - middle (conflict), Act 3 - ending (resolution)). Alternatively, a scenario in production may be a scenario in which the first act or part of the second act has been written.

[0021] (Control unit 12) The control unit 12 functions as an arithmetic processing unit and a control device, and controls the overall operation of the information processing device 1 in accordance with various programs. The control unit 12 is realized by electronic circuits such as a CPU (Central Processing Unit) or a microprocessor. The control unit 12 may also include a ROM (Read Only Memory) that stores the programs to be used, arithmetic parameters, etc., and a RAM (Random Access Memory) that temporarily stores parameters that change as appropriate.

[0022] Furthermore, the control unit 12 according to this embodiment also functions as an element extraction unit 121, an output information generation unit 122, and an output control unit 123. The element extraction unit 121 has a function of extracting story elements from scenario information. Even if elements are not input from the input unit 11, it is possible to extract elements from the scenario information. The element extraction unit 121 analyzes the scenario information (text data) to extract elements such as characters (relationships), setting (period, location), and props (stage props, small props). For example, natural language processing may be used for the analysis. The element extraction unit 121 can extract the above-mentioned elements by performing natural language processing (morphological analysis, syntactic analysis, anaphora analysis, etc.) on descriptions such as headings, stage directions, and lines included in the scenario information.

[0023] The output information generation unit 122 generates output information to be presented to the user based on the element information. The output information to be presented to the user is information to support the creation of a better scenario. More specifically, for example, the output information generation unit 122 estimates a predetermined value that will be generated by the content to be generated based on information about the set elements (setting elements) that make up the story, compares the estimated value with a target value input from the input unit 11, and generates information regarding the correction of the setting elements based on the comparison result. The information regarding the correction of the setting elements is information regarding the addition or deletion of setting elements. Based on the comparison result between the estimated value and the target value, the output information generation unit 122 determines whether to add or delete setting elements to bring the content closer to the target value. For example, the output information generation unit 122 estimates the duration of the content when visualized based on the setting element information. Then, the output information generation unit 122 compares the estimated duration with the duration input as the target value and determines whether to add or delete setting elements to bring the content closer to the target duration based on the comparison result. For example, if the estimated time length is longer than the target time length, the output information generating unit 122 determines which setting elements to delete, and if it is shorter, determines which setting elements to add.

[0024] As another example of information for supporting better scenario creation, the output information generation unit 122 may generate information for generating simulation video (so-called previsualization) for imagining the finished state before content production, such as actual filming or CG production, begins. Such simulation video may be created using a simple CG (Computer Graphics) model. Furthermore, the simulation video may be referenced when determining camerawork, character placement, VFX (Visual Effects), editing, and the like in advance. The output information generation unit 122 may generate information necessary for processing when visualizing a scenario with simulation video by analyzing scenario information (text data). Specifically, the output information generation unit 122 performs natural language processing (morphological analysis, syntactic analysis, anaphora analysis, etc.) on the scenario information to extract elements for visualization (elements constituting a story; components). Similar to the setting elements described above, components may include characters, settings (time period, location), and props (stage props, small props) that constitute the story. The output information generation unit 122 may perform visualization processing (automatic generation of a simulation video using simple CG) based on the extracted components and generate a simulation video as output information. In the visualization processing, images corresponding to the components are searched for and automatically generated, and visualized (visualized) for each scene. In this case, the output information generation unit 122 may separate components into those that can be modified and operated by the user and those that are automatically generated, and acquire images corresponding to each. For example, the output information generation unit 122 may use pre-prepared 3DCG assets (model data) for components that can be modified and operated by the user. Such distinctions between components may be determined, for example, based on the importance of the component. Components with an importance higher than a threshold (or determined to be important) are considered "important elements" and are treated as modifiable and operable components. This allows the user to modify and operate the appearance, position, etc. of the modifiable and operable 3DCG included in the simulation video visualized for each scene via the input unit 11.

[0025] The output control unit 123 controls the output unit 13 to output the output information generated by the output information generation unit 122 .

[0026] (Output section 13) The output unit 13 has a function of outputting information. For example, the output unit 13 may be a display unit, an audio output unit, a projector, a communication unit, or the like.

[0027] (Storage unit 14) The storage unit 14 is realized by a ROM (Read Only Memory) that stores programs and calculation parameters used in the processing of the control unit 12, and a RAM (Random Access Memory) that temporarily stores parameters that change as appropriate.

[0028] The basic configuration of the information processing device 1 according to this embodiment has been described above. Note that the basic configuration of the information processing device 1 is not limited to the example shown in FIG. 1. For example, the information processing device 1 may be realized by a plurality of devices. Specifically, for example, the control unit 12 and the storage unit 14 may be provided in a server, and the input unit 11 and the output unit 13 may be provided in a user terminal (such as a PC, a smartphone, or an HMD (Head Mounted Display)).

[0029] Next, the scenario creation support according to this embodiment will be described in more detail.

[0030] <<3. First Embodiment>> In the first embodiment, information (proposal content) relating to the correction of setting elements is output to support scenario creation. The configuration and operation processing of the first embodiment will be described below in order.

[0031] <3-1.Configuration example> Fig. 2 is a block diagram showing an example of the configuration of the control unit 12A according to the first embodiment. As shown in Fig. 2, the control unit 12A has a setting element extraction unit 121A, an output information generation unit 122A, an output control unit 123A, and a tagging processing unit 124. The output information generation unit 122A also functions as an estimation unit 1221, a comparison unit 1222, a correction information generation unit 1223, and a display screen generation unit 1224. For the sake of explanation, the example shown in Fig. 2 also shows a past work knowledge DB 141, a past work setting element DB 142, and a setting element change history DB 143, which are DBs (databases) included in the storage unit 14.

[0032] (Setting element extraction unit 121A) The setting element extraction unit 121A extracts information about setting elements from information about a scenario. For example, the setting element extraction unit 121A performs natural language processing on the scenario information (text data) of a given past work (e.g., a movie) to extract information about setting elements such as the characters, time period, location, props / stage equipment, etc. in the past work, and stores the information in the past work setting element DB 142. Furthermore, when scenario information currently being produced is input, the setting element extraction unit 121A similarly performs natural language processing on the scenario information to extract information about setting elements and outputs the information to the estimation unit 1221.

[0033] For example, to extract a character, the setting element extraction unit 121A extracts information such as the character's name, a list of the character's lines, a list of the character's actions (obtained from verbs), and character settings (relationship with the protagonist). The setting element extraction unit 121A also assigns a character ID and importance to the extracted character. The importance is the character's importance in the story, and can be determined, for example, from the amount of lines, the number of scenes in which the character appears, character settings, etc. The setting element extraction unit 121A determines a character with a high level of importance as a main character (Main), and a character with a low level of importance as a supporting character (Sub). There are no particular limitations on the criteria for this determination.

[0034] (Estimation Department 1221) The estimation unit 1221 estimates a value resulting from content generated based on the setting elements, based on the setting element information. Examples of values ​​resulting from the content include the duration of time when visualized, the filming cost, or profits. The setting element information may be information extracted by the setting element extraction unit 121A from information about a scenario being created, or may be information about setting elements input by a user. The setting element extraction unit 121A may perform estimation using only the setting element information input by the user (e.g., character names and character settings), or, if scenario information is also input, may also perform estimation using setting element information extracted from the scenario information by natural language processing (e.g., a list of lines and a list of actions). The more information input, the higher the accuracy of the estimation.

[0035] The estimation may be performed, for example, using the results of learning from past works. Since the duration, filming costs, profits, etc. of past works are known, various values ​​for each setting element can be calculated based on the information on the setting elements of the past works and this known information. Such learning may be performed in advance, and the learning results may be stored in the past work setting element DB 142. Alternatively, the estimation unit 1221 may perform estimation (learning) based on the information on the setting elements of past works stored in the past work setting element DB 142 and information on the duration, filming costs, profits, etc. of the past works stored in the past work knowledge DB 141. Based on the information on the setting elements of the current work, the estimation unit 1221 estimates values ​​that may be generated by the information on the setting elements of the current work from values ​​associated with the information on the same setting elements in past works. More specific details of the estimation according to this embodiment will be described later (see FIGS. 4 to 6).

[0036] (Comparator 1222) The comparison unit 1222 compares the estimated value estimated (calculated) by the estimation unit 1221 with the target value. The target value may be input by the user or may be set in advance. For example, if the estimation unit 1221 estimates the time length to be "8 minutes" based on the information on the setting elements of the current work and the target value is "10 minutes," the comparison unit 1222 outputs a comparison result such as "2 minutes short."

[0037] (Correction information generation unit 1223) The correction information generation unit 1223 generates information (correction information) regarding correction of setting elements, such as adding or deleting setting elements, based on the comparison result by the comparison unit 1222. For example, if the comparison result shows that the estimated value is less than the target value, the correction information generation unit 1223 determines to add a setting element, and if the estimated value exceeds the target value, the correction information generation unit 1223 determines to delete a setting element. The correction information generation unit 1223 can determine which setting elements to add or delete based on information about the setting elements (such as the time length for each setting element) obtained from the past work setting element DB 142 and the amount of profit from past works obtained from the past work knowledge DB 141. More specific details of correction information generation according to this embodiment will be described later.

[0038] (Display screen generation unit 1224) The display screen generation unit 1224 generates a display screen for presenting the correction information generated by the correction information generation unit 1223 to the user, and outputs the display screen to the output control unit 123A. For example, the display screen generation unit 1224 may generate a screen on a character setting screen (an example of an input screen for setting elements) that shows a comparison result between a time length estimated from the current character setting and a target value, and displays a sentence proposing a setting element to be added / deleted (for example, a new character) as an "AI suggestion" (see FIG. 7).

[0039] Furthermore, if a scenario is currently being created (if information about the currently being created scenario has been input), the display screen generation unit 1224 may generate a screen showing which parts of the scenario text will be affected when a user adopts a setting element correction. The display screen generation unit 1224 updates the display screen as needed in response to user operations, such as when a setting element change instruction (operational input to adopt the proposed setting element correction) is made. Here, particularly when the deletion of a setting element is adopted, consistency may be lost if related sentences are not deleted. When a setting element is changed (for example, when the user adopts the proposed deletion of a setting element), the display screen generation unit 1224 performs a text search process to identify sentences that reference the setting element (for example, sentences that contain the name of the setting element), and generates a display screen that clearly indicates this to the user and displays a sentence that prompts the user to delete or correct the setting element. Note that the same setting element may be expressed using different words in the scenario text. For example, the same person may be referred to by name and by their job title. Therefore, the display screen generation unit 1224 may perform a text search process by referring to the processing result (see FIG. 8) by the tagging processing unit 124 that associates the scenario text with the setting elements. More specifically, the display screen generation unit 1224 performs a text search process for the setting elements to be deleted on the scenario text that has been tagged, and identifies the sentences associated with the setting elements (portions to which the tags of the setting elements have been added) (see FIG. 9). The user can arbitrarily set whether the specified range is to be per sentence or per paragraph.

[0040] It is also possible that the user may have adopted a proposed correction (addition / deletion) to a setting element but later wish to revert it to its original state, or may have rejected the proposed correction but later wish to revert it. Therefore, in this embodiment, the correction information (proposal content) generated by the correction information generation unit 1223 and the changes made by the user to the scenario text and setting elements (such as whether the proposal was adopted, addition, deletion, and content before the change) are stored as history in the setting element change history DB 143. The display screen generation unit 1224 may refer to the information stored in the setting element change history DB 143 and display the proposed content adopted so far and the corresponding edit content in a part of the display screen, for example, using a card-type UI (user interface) (see FIG. 10 ). This allows the user, when wanting to revert a change to each setting element, to select a “rollback” button, which allows the control unit 12A to revert the scenario to the state before the change (rollback; backward reversion). The display screen generation unit 1224 may also generate a screen displaying proposed but rejected proposal content in a card-type UI. The user can change the scenario settings at any time by selecting the "Adopt" button on the card-type UI.

[0041] (Tagging processing unit 124) The tagging processor 124 performs a process of associating (tagging) setting elements with the main scenario text based on input information about the scenario being created and information about setting elements extracted from the scenario. As described above, for example, there are cases where the same person is listed by name (the "name" of the setting element) and also by the person's job title (the "character setting" of the setting element), but by tagging the words in the main scenario text with the tagging processor 124, it becomes clear that one setting element corresponds to one or more different words that appear in the main scenario text. A specific example of tagging will be described later with reference to FIG. 8.

[0042] (Output control unit 123A) The output control unit 123A controls the display of the display screen generated by the display screen generation unit 1224 on, for example, a display unit (an example of the output unit 13).

[0043] The above has described an example of the configuration of the control unit 12A according to the first embodiment. Note that the configuration shown in Fig. 2 is just an example, and the present embodiment is not limited to this.

[0044] <3-2. Operation processing> FIG. 3 is a flowchart showing an example of the flow of the operation process according to the first embodiment.

[0045] 3, first, the control unit 12A of the information processing device 1 receives an input of a target value from the input unit 11 (step S103). In this flow, as an example, a case will be described in which the "duration when a scenario is visualized" is input as the target value.

[0046] Next, the setting element extraction unit 121A of the control unit 12A performs natural language processing on the input information (text data) of the scenario being created (current work) to extract information on setting elements such as characters (character names, character settings, list of lines, list of actions, etc.), time period, and location (step S106). At this time, the setting element extraction unit 121A may calculate the importance of each setting element based on the content of the scenario. The importance is calculated (determined) based on the character settings (protagonist, protagonist's lover, protagonist's rival, etc.), the number of times the character appears in the scenario, whether the character has a lot of dialogue, whether the setting element is described in detail, etc. The setting element extraction unit 121A assigns a determination result to each setting element, such as "Main" if the importance is higher (than a threshold value) or "Sub" if the importance is lower (than a threshold value).

[0047] The control unit 12A also accepts input of setting elements (for the current work) from the input unit 11 (step S109). In this embodiment, the user may input only the setting elements, or may input information about the scenario being created and have the information processing device 1 extract the setting elements. The user may also input both information about the scenario being created and the setting elements.

[0048] Next, the estimation unit 1221 estimates the time length when the current work is visualized based on the information on the acquired setting elements (step S112). Various estimation methods are conceivable, but in this flow, as an example, a case where this is performed by learning the setting elements of past works and the time length of the images of the past works will be described. More specifically, the estimation unit 1221 calculates the time length of each setting element based on the information on the setting elements extracted from the scenario of the past work and the time length of the past work. Note that this calculation may be performed in advance. The following description will be made with reference to FIG. 4.

[0049] FIG. 4 is a diagram illustrating learning of the duration of each setting element of a past work according to the first embodiment. First, since the duration of the past work is known, the formula shown in FIG. 4 is established based on the number of words spoken by each character and the type of action taken by each character. That is, based on the line list and action list (list of verbs) for each character obtained by extracting the setting elements, the time (playback time) required to visualize these is estimated. The line playback time may be estimated using statistics of the average speaking speed in each language. Furthermore, the action playback time may be estimated by searching for the corresponding motion in a database (not shown) that collects motion data separately and using the playback time of that motion. For motions that do not exist in the database, the average playback time of all motion data may be used. Furthermore, due to differences in the speed of each movement and the time between actions, simple addition will not match the actual duration when visualized. Therefore, a correction constant is prepared and multiplied to calculate a correction constant that matches the duration of the work (100 minutes in the example shown in FIG. 4). While such correction constants vary depending on various conditions, here, as an example, an average correction constant for each movie genre is calculated. The estimation unit 1221 can calculate the average correction constant for each movie genre by executing the calculation formula shown in Fig. 4 for a large number of past works. Note that there is a tendency for the average correction constant for each movie genre; for example, this value is large for horror movies and small for action movies.

[0050] FIG. 5 is a diagram illustrating the estimation of the duration of each setting element of the current work according to the first embodiment. The estimation unit 1221 first allocates the duration of the entire work to each character based on the average correction constant calculated for each movie genre, the extracted dialogue list for each character, and the extracted action list for each character. For example, in the example shown in FIG. 5, the estimation unit 1221 estimates the duration of each setting element other than dialogue and action based on the duration of each character. Specifically, as shown in FIG. 5, the estimation unit 1221 estimates the duration of each setting element other than dialogue and action (average) based on the duration of each character. Specifically, as shown in FIG. 5, the estimation unit 1221 estimates the duration for each "importance" and "character setting" (average). This estimation process of the duration for each "importance" and "character setting" can be performed for all past works managed by the information processing device 1. A database containing the results of this estimation process for a large number of past works may be prepared and used by the information processing device 1. The information processing device 1 may also obtain information about past works (such as scenario information and video duration) from the Internet, make estimates, and store the results in a database. Note that this estimation process is not limited to characters, but can also be performed on other setting elements, such as locations and props / stage equipment. Since these setting elements do not have lines or actions, the duration of each setting element can be estimated using the number of scenes in which they appear. The above-described estimation process can also be applied to machine learning using an RNN (Recurrent Neural Network).

[0051] The estimation unit 1221 then uses the learning results to estimate the duration of each setting element in the current work, and calculates the sum of the estimated values ​​for each setting element as the estimated duration of the current work. In the example shown in the bottom of Figure 5, the duration of "BOB," a setting element in the current work, is estimated. In this case, if there are multiple setting element items that can be used (e.g., "Importance" and "Character Setting") and different estimated values ​​are available for each (e.g., 3.5 minutes for "Importance: Sub," 4 minutes for "Character Setting: Father"), one of them may be selected according to a predetermined rule. For example, the number of times the same setting element item has been used in past works may be calculated, and the setting element item with the smallest number of times may be used. In the example shown in the bottom of Figure 5, between "Importance: Sub" and "Character Setting: Father," for example, the element "father" has appeared less frequently in past works, so its estimated value of 4 minutes may be selected. This is because items that appear more frequently are more likely to have larger deviations.

[0052] The estimation unit 1221 can also switch the learning data of past works used in the estimation process as appropriate. Since the scenarios of past works are created by various creators, there are differences in writing style and the level of detail of the descriptions. For example, if a noise case occurs as described below, it may become impossible to perform an accurate estimation. An example of a noise case is when "the extraction of setting elements does not go well due to the writing style." For example, examples include cases where the setting element is divided into multiple sentences and the subject is omitted (example sentence: "There is a desk, with monitors, and a chair. But apparently no one inside."), cases where the object is omitted (example sentence: "JAY exits." ("JAY exits the door."), cases where it is written using a pronoun (example sentence: "There is a desk, with monitors, and a chair." + "It is JAY's."), and cases where it is modified at a structurally distant point (example sentence: "There is a desk, with monitors, and a chair. . . . The desk is JAY's."). It is also possible that even if the setting element is properly extracted, it is not written in the scenario (the information about the extracted setting element is less than expected). Therefore, in this embodiment, the likelihood of the setting element extraction result is calculated for each work, and the learning data of past works to be used may be switched based on this. The calculation of the likelihood may be performed by the estimation unit 1221, or may be performed in advance by an external device.

[0053] Here, an example of calculating the likelihood of a work will be described. For example, the number of sections that fall into the above-mentioned "noise cases" is counted throughout the entire scenario text, and the normalized result is calculated by dividing by the number of words in the entire scenario text. Specifically, likelihoods are calculated according to the following writing style and level of detail, and the result is multiplied to obtain the final likelihood of the work. (1) Likelihood by writing style: sum of the following items · The sentence is divided into multiple sentences and the subject is omitted → Syntactic analysis counts the places where the subject is missing. · The object is omitted → Count the places where the subject is missing through syntactic analysis. · Written using pronouns → Count the places where the subject is missing through syntactic analysis. · Modified at structurally distant points → Count the number of times that the same noun appears in the same paragraph at least two sentences apart. (2) Likelihood based on the level of detail of the description: Add the following value - Coverage rate of the information extracted from the setting elements → Average extraction rate of item information for each setting element

[0054] When estimating a predetermined value for the current work (for example, the duration when visualized), the estimation unit 1221 can switch the learning data to be used by referring to the likelihood of past works. Specifically, for example, the estimation unit 1221 calculates the likelihood of the input scenario currently being produced, compares the calculated likelihood with the likelihood of past works, and uses learning data from past works with similar values, thereby enabling relatively accurate estimation. In the case of a creator who writes scenarios with relatively omissions of subjects and objects, noise due to differences in the amount of description and differences in extraction accuracy can be reduced by comparing the scenarios with scenarios from past works that often have similar omissions.

[0055] Next, the comparison unit 1222 compares the estimated value with the target value (step S115), and determines whether the difference between the estimated value and the target value is equal to or greater than a specified value (step S121).

[0056] If the difference is not greater than the specified value (step S121 / No), the output information generation unit 122A generates a display screen displaying the estimated value (the estimated time length when the current work is visualized) using the display screen generation unit 1224, and displays it on the display unit using the output control unit 123A (step S121).

[0057] On the other hand, if there is a difference equal to or greater than the specified value (step S121 / Yes), the output information generation unit 122A uses information on the setting elements of the past work to generate correction information for the setting elements (determine setting elements to propose addition or deletion) using the correction information generation unit 1223 (step S124). Note that the level of discrepancy at which this generation process is executed can be adjusted (by the user) by changing the specified value. If the estimated value is less than the target value, the correction information generation unit 1223 determines setting elements to add, and if the estimated value exceeds the target value, it determines setting elements to delete from the setting elements of the current work.

[0058] FIG. 6 is a diagram illustrating the case where setting elements to be added are selected from past works. The correction information generation unit 1223 utilizes the learning data of past works used when the estimation unit 1221 performs the estimation process described above to select setting elements that match the numerical values ​​to be increased. That is, as shown in FIG. 6, the correction information generation unit 1223 selects setting elements whose estimated values ​​fill the gap with the target value from the estimated values ​​(e.g., duration) for each setting element obtained by learning from past works. However, this condition alone results in a large number of setting elements to be selected, and there is also a risk that setting elements that overlap with setting elements of the current work will be selected. Therefore, the correction information generation unit 1223 may determine setting elements suitable for addition to the current work using the following criteria (a) to (c). (a) Similarity between this work and previous works (priority is given to works whose content is similar to the scenario currently being produced) (b) The inverse of the degree of similarity between the setting elements of this work and those of previous works (priority is given to setting elements that are least similar to the setting currently being created). (c) The amount of revenue (e.g., box office revenue) of the work (priority selection is given to popular past works)

[0059] The similarity between works may be, for example, a value calculated by calculating the distance between vectors, with each setting element in the work as an element. In practice, the above three measures may be calculated, normalized, and multiplied together to obtain an evaluation score, with the result output as candidates to be suggested for addition in descending order of evaluation score. The evaluation score is calculated for all setting elements of similar past works. Note that the weighting of each measure during normalization may be variable, allowing the user to select. In this way, the correction information generation unit 1223 can determine, as candidates for additional setting elements, setting elements from past works that are similar to the scenario currently being created (e.g., similar genre, character settings, setting, etc.), do not overlap with setting elements of the current work, and are as likely to be popular (profitable) as possible.

[0060] In the example shown in FIG. 6, previous work B is more similar to the scenario currently being created than previous work A (because previous work B contains the common setting element "character setting: uncle"), and within previous work B, a setting element "EMILY" is selected that is not similar to the information on the setting elements of the current work (because it contains the uncommon "character setting: mother"). In addition, the information on the time length of the setting element "EMILY", "3 minutes", is used to create a proposal, which will be described later. The action of the setting element "EMILY" may also be included in the proposal content.

[0061] On the other hand, if the estimated value based on the information on the setting elements of the current work exceeds the target value and it is better to delete some of the setting elements of the current work, the correction information generation unit 1223 may refer to the estimated value for each setting element of the current work (see FIG. 5) and select a setting element with an optimal value to eliminate the difference from the target value. Also in this case, since it is expected that there will be multiple candidates, the correction information generation unit 1223 may determine setting elements suitable for deletion using the following criteria (A) to (C). (A) Similarity between this work and previous works (priority will be given to works whose content is similar to the scenario currently being produced) (B) The inverse of the degree of similarity between the setting elements of this work and those of previous works (priority is given to setting elements that are least similar to the setting currently being created). (C) The work's revenue (for example, box office revenue) is low (setting elements that are the same (similar) to unpopular previous works are prioritized for deletion).

[0062] The correction information generating unit 1223 normalizes and multiplies these values, and selects them in order as candidates for deletion.

[0063] In the example described above, characters were used as an example for selecting setting elements to add / delete, but this embodiment is not limited to this, and similar processing can be performed for other setting elements such as the setting (time period, location) and props / stage props to select candidates for addition / deletion for each.

[0064] Furthermore, the importance of the setting elements may be taken into consideration when calculating the evaluation score that determines the priority order of addition / deletion candidates. For example, among addition candidates, setting elements with higher importance may be prioritized, and among deletion candidates, setting elements with lower importance may be prioritized.

[0065] The correction information generation unit 1223 can also switch the learning data of past works used when determining candidates to add / delete, depending on the likelihood, as in the case of the estimation process. The calculation of likelihood is as described above. In order to search for candidates in a wider range, the correction information generation unit 1223 may lower the likelihood threshold and use learning data of more past works to determine candidates to add / delete.

[0066] Next, the correction information generating unit 1223 generates a sentence for proposing the addition or deletion of a setting element (step S127). The correction information generating unit 1223 may generate the sentence by applying the selected addition candidate / deletion candidate to a predetermined sentence template, for example.

[0067] Next, the control unit 12A controls the output control unit 123A to display the display screen generated by the display screen generation unit 1224, which proposes adding or deleting a setting element, on the display unit (step S130). FIG. 7 is a diagram showing an example of a display of a setting element change proposal according to the first embodiment. A screen 410 shown in FIG. 7 is a screen displayed on the display unit of the information processing device 1. The screen 410 shows a state in which a person correlation tab 411 is selected as the input of a setting element, and an input screen 412 of person correlation is displayed. The method of inputting person correlation is not particularly limited. For example, the user may create a correlation diagram by entering text or selecting an icon, or may select a correlation diagram template prepared in advance and modify the template. Furthermore, a target value (e.g., a time length) may also be input on the input screen 412. A screen 413 showing the change proposal is displayed on the right side of the display screen 410. The screen 413 may be a card-type UI. Such a change proposal may be displayed immediately each time the person correlation on the input screen 412 is changed. The user decides whether to adopt the proposal. This is because, taking into consideration the preferences of the user (creator) and the consistency of the final overall scenario, it is considered preferable for a human to make the judgment. Furthermore, multiple proposed changes may be made. These are displayed one by one in a card-type UI, and each has a button for whether to adopt it or not. The user can arbitrarily select from one or more proposals (card-type UI). The control unit 12A reflects the adopted proposal in the person correlation on the input screen 412, and performs the estimation process again (of the time length of the current work (being input)).

[0068] Next, if the user adopts the suggestion (step S133 / Yes), the display screen generation unit 1224 searches for the extent to which adopting the suggestion will affect the scenario (step S136) and displays the extent to which the scenario will be affected (step S139). This is because, if a scenario in the process of being created has been input, changing the setting elements may result in inconsistency with the content of the scenario. As described above, the tagging processing unit 124 associates the setting elements with words in the main text of the scenario, and the display screen generation unit 1224 performs a text search process by referring to the processing results, and identifies the extent to which adopting the suggestion will have an impact.

[0069] The tagging process can be performed every time the scenario text is updated through editing. FIG. 8 is a diagram illustrating the tagging process according to the first embodiment. As shown in FIG. 8, for example, words (here, "Suzuki" and "boss") present in the information of one extracted setting element are extracted from the scenario text and replaced with a predetermined tag name (here, "@boss"). This associates the scenario text with the setting element, and even if the word "boss" simply appears in another part of the scenario text, it can be determined that this refers to "Suzuki."

[0070] The display screen generation unit 1224 searches for tags of setting elements that have been selected for deletion in the tagged scenario text, extracts sentences or paragraphs where the tags exist, and generates a screen that presents the user with the range of influence. FIG. 9 is a diagram showing an example of a screen displaying the range of influence on the scenario caused by the deletion of a setting element according to the first embodiment. As shown in FIG. 9, a screen 420 shows a state in which a scenario tab 421 is selected, and a range 422 of influence on the scenario caused by the deletion of a setting element (here, the character "boss") is displayed. The range 422 of influence on the scenario is displayed, for example, in highlighting. Note that the user can change whether the range 422 of influence on the scenario is displayed in sentence units or paragraph units by setting. Furthermore, a screen 423 showing deletion suggestions regarding the impact on the scenario can be displayed using a card-type UI, as shown in FIG. 9. To delete all of the range 422 of influence on the scenario, the user selects the "Delete" button on the screen 423.

[0071] As described above, it is also possible that a user may adopt a suggestion but later wish to revert to the original state. In this embodiment, the proposed content and the changes made by the user in response to the suggestion may be stored as a history in the setting element change history DB 143, and the change history may be listed in a card-type UI as shown in FIG. 10. FIG. 10 is a diagram showing an example of a screen displaying the history of changes made in response to the suggestion according to the first embodiment. If the user wishes to revert changes made to each setting element, the user can select the "Roll back" button to return the scenario body to the state before the changes.

[0072] On the other hand, if the user does not adopt the suggestion (step S133 / No), the output information generating unit 122A stores information about the suggestion in the storage unit 14 (step S142).

[0073] <3-3. Display screen example> FIG. 11 is a diagram illustrating an example of screen transitions in scenario creation support according to the first embodiment. As shown in FIG. 11, for example, first, the title and outline of the work to be created are entered on screen 450, and then a genre is selected on screen 460. An example of screen 450 (title and outline determination screen) is shown in FIG. 12. As shown in FIG. 12, on screen 450, the work title, image (picture), logline (a sentence summarizing the story content in one line), the target time length when visualized, estimated shooting period, etc. are entered. A screen for managing alternative ideas may be provided as part of screen 450. The screen for managing alternative ideas may also display messages from other users, such as team members, and allow users to brainstorm ideas together. An example of screen 460 (genre selection screen) is shown in FIG. 13. As shown in FIG. 13, on screen 460, the story type (genre) is selected from templates.

[0074] Next, as shown in FIG. 11 , settings for the entire work are input on each of the screens: a character setting screen 470, a location setting screen 480, and a details (stage props, props, etc.) setting screen 490. On each setting screen, setting element information (learning data) extracted from scenario information of previous works may be used to appropriately suggest the addition / deletion of setting elements. An example of the character setting screen 470 is shown in FIG. 14 , an example of the location setting screen 480 is shown in FIG. 15 , and an example of the details (stage props, props, etc.) setting screen 490 is shown in FIG. 16 . As shown in each figure, on each setting screen, suggestions for adding / deleting setting elements are displayed with an "AI" icon. When the user adds / deletes setting elements according to the suggestions, the content of the suggestions is updated.

[0075] Once the settings for the entire work are input, the actions (events) for each scene are edited on a beat sheet editing screen 500, as shown in FIG. 11 . Here, too, using setting element information (learning data) extracted from scenario information of past works, suggestions for adding / deleting setting element information (here, actions) can be made as appropriate. FIG. 17 is a diagram showing an example of the beat sheet editing screen 500. As shown in FIG. 17 , on the beat sheet editing screen 500, actions (events) are arranged in the order of scene development for each act that makes up a story, for example. Note that a screen for managing alternative ideas may also be provided in part of the screen 500. The screen for managing alternative ideas can display messages from other users, such as team members, and enable users to brainstorm ideas together. Furthermore, on the screen for managing alternative ideas, suggestions for adding / deleting setting elements are also displayed with an "AI" icon.

[0076] Next, as shown in FIG. 11, the specific content of the plot is determined on an editing screen 510 for the developing plot. FIG. 18 is a diagram showing an example of the editing screen 510 for the developing plot. As shown in FIG. 18, on the editing screen 510 for the developing plot, the plots are arranged in the order of scene development. Note that a screen for managing alternative ideas may also be provided in part of the screen 510. On the screen for managing alternative ideas, messages from other users such as team members can be displayed, and ideas can be exchanged with other users. Furthermore, on the screen for managing alternative ideas, suggestions for adding / deleting setting elements are also displayed with an "AI" icon.

[0077] Then, based on the above content, the user finally creates a scenario on the scenario editing screen 520. Fig. 19 shows an example of the scenario editing screen 520. As shown in Fig. 19, on the scenario editing screen 520, the main text of the scenario, such as headings, stage directions, and dialogue, is input.

[0078] The above describes an example of screen transitions in the scenario creation process after determining the overall setting and plot of the work, and suggestions for setting elements on each screen. Note that the above-described screen transitions are merely examples, and the present embodiment is not limited to these.

[0079] For example, the information processing device 1 may extract setting elements from a scenario in production that is input on the scenario editing screen 520, and may appropriately propose changes to the setting elements in accordance with the target values.

[0080] Furthermore, in the screen transition shown in FIG. 11 , since there are few setting elements entered in the initial setting screens, such as the character setting screen 470, the content of the additional proposal is likely to not match the content being created. Therefore, in this embodiment, if the number of input setting elements is less than a certain number, or if the difference from the target value is greater than a threshold, the additional proposed element is presented only as a reference. Note that in the screen transition shown in FIG. 11 , the title and genre are initially required input fields, and their word vectors are set as initial setting elements, so there is no case where there are no setting elements. Furthermore, when suggesting additional setting elements for reference, multiple setting elements may be proposed as a single block rather than presented one by one. FIG. 20 is a diagram illustrating a case in which character correlations are proposed for reference in units of correlation diagrams according to the first embodiment. As shown in the upper part of FIG. 20 , if, for example, only one setting element has been entered in the character setting screen, the information processing device 1 makes a proposal for each correlation diagram. If the user adopts the proposal, all setting elements being input are replaced with the proposed correlation diagram, as shown in the lower part of FIG. 20 . Furthermore, the setting element being input is saved as an alternative.

[0081] <3-4. Application Examples> The estimation by the estimation unit 1221 is not limited to estimating the duration of the visualization, but can also estimate, for example, the filming cost. The filming cost may refer to the time required for filming and CG production, or may be an expense based on time and labor costs. Generally, the more characters, locations, lines of dialogue, and actions of characters, the higher the filming cost. The operation process for proposing setting elements based on the filming cost (generating correction information for setting elements, such as addition / deletion) is performed in the same manner as the operation process shown in FIG. 3. The difference is that in step S103, the "filming cost" is input as a target value. Also, in step S112, the estimation unit 1221 estimates the filming cost for filming the current work. Specifically, the estimation is performed by referring to learning data from past works. Specifically, by learning using information on setting elements extracted from the scenarios of past works and data on the filming cost (filming time) of each scene obtained as knowledge from past works, the extent to which each setting element affects the filming cost is roughly estimated. Then, in step S115, the target value is compared with the estimated value, and if in step S118 the difference between the estimated filming cost of the current production and the target value is equal to or greater than a specified value, setting elements to be added or deleted are determined in step S124. For example, if the estimated filming cost of the current production is greater than the target value, a proposal to delete setting elements (such as deleting characters or locations, or reducing filming time) is made. Note that reducing filming time is expected to reduce filming costs as well.

[0082] The estimation unit 1221 can also estimate various other values ​​by using various information about past works. For example, the estimation unit 1221 can also estimate box office revenue based on setting elements. The estimation unit 1221 learns from setting element information extracted from the scenarios of past works and box office revenue data obtained as knowledge of past works, thereby estimating the extent to which each setting element affects box office revenue. In the case of box office revenue, it is considered that each line of dialogue or action has little impact on box office revenue, and therefore the box office revenue forecast for each setting element is calculated without performing the approximation processing or correction constant processing shown in FIG. 4.

[0083] Furthermore, the estimation unit 1221 can also perform popularity estimation processing using positive / negative judgments of word-of-mouth reviews of past works, time-series data on the emotional movements of audiences and characters when watching a video work, known as an emotional curve, and the like.

[0084] <<4. Second Embodiment>> In the second embodiment, to support scenario creation, information for generating simulation footage (so-called previs, which visualizes a scenario using simple CG) is generated from scenario information (to visualize the scenario). Furthermore, in the second embodiment, it is also possible to distinguish and visualize automatically generated components from components that can be modified or manipulated. Generally, the production process for videos such as movies and commercials follows the steps of "planning → shooting → editing → finishing," and previs (simulation footage) can be created between planning and shooting. In this embodiment, components such as locations, characters, props, and stage equipment (synonymous with the "setting elements" in the first embodiment) and detailed information such as their movements (referred to as attribute information in this embodiment) are extracted from the scenario (script), and a previs is automatically generated (the scenario is visualized). The configuration and operation process of the second embodiment will be described below.

[0085] <4-1.Configuration example> Fig. 21 is a block diagram showing an example of the configuration of a control unit 12B according to the second embodiment. As shown in Fig. 21, the control unit 12B has a component extraction unit 121B, an output information generation unit 122B, an output control unit 123B, a component estimation unit 126, an importance determination unit 127, and a label assignment unit 128. In addition, the output information generation unit 122B functions as a performance proposal unit 1226, a command generation unit 1227, and a visualization processing unit 1228. For the sake of explanation, the example shown in Fig. 21 also shows a current work component DB 145, a past work component DB 146, a general knowledge DB 147, and a command DB 148, which are DBs (databases) included in the storage unit 14.

[0086] (Component extraction unit 121B) The component extraction unit 121B extracts components for visualization from the input information (text data) of the scenario in production and stores the extracted component information in the current work component DB 145. The component extraction unit 121B performs natural language processing such as morphological analysis, syntactic analysis, and anaphora analysis on the scenario information (text data) to extract the components to be visualized. The control unit 12B formats the input scenario information for analysis and passes it to the component extraction unit 121B as a text file. The component extraction unit 121B first extracts the "lines" of the characters, "stage directions" that are sentences that instruct actions and direction, and "headlines" that explain the location and time period. From these descriptions, the component extraction unit 121B then extracts attribute information of scene-independent components (overall metadata), such as characters, locations, and props, in a format that follows predetermined rules (see FIGS. 25 to 27). Furthermore, the component extraction unit 121B extracts attribute information of scene-dependent components (time-series metadata) such as the movements of characters and changes in props / stage equipment from the "stage directions" and "header descriptions" (see FIGS. 30 and 31). Details of the component extraction process will be described later with reference to FIG. 23.

[0087] Information on the components extracted by the component extraction unit 121B (overall metadata, time-series metadata) is stored in the current work component DB 145. Furthermore, after various processes according to this embodiment are completed, the control unit 12B may ultimately transfer only the "overall metadata" from the current work component DB 145 to the past work component DB 146. The current work component DB 145 stores information on components specific to the input text, which can be modified by the user as appropriate. Furthermore, the past work component DB 146 stores information on components of past works that have already been analyzed.

[0088] (Component element estimation unit 126) The component estimation unit 126 estimates attribute information in a format that follows a predetermined rule in order to supplement information (attribute information) of components that could not be extracted by the component extraction unit 121B. The component estimation unit 126 may use a machine learning model to estimate missing information from scenario information and information on already extracted components. The estimated attribute information is stored in the current work component DB 145. Details of the component estimation process will be described later with reference to FIG. 24.

[0089] (Importance determination unit 127 and label assignment unit 128) The importance determination unit 127 determines the importance of the extracted or estimated constituent elements within the story. The importance determination may be performed simultaneously with the extraction process by the constituent element extraction unit 121B, or may be performed after the extraction process and estimation process are completed. Specifically, the importance determination unit 127 determines the importance of each constituent element in the overall scenario (work) and the importance of each scene. The importance determination is calculated (determined) based on the number of times the constituent element appears, the detail of the drawing, the amount of conversation, etc. Details of the constituent element importance determination process will be described later with reference to FIG. 32.

[0090] The labeling unit 128 assigns a label of automatic / manual (user-modifiable / operable) to each component element according to the determination result by the importance determination unit 127. The labeling unit 128 assigns a label of "manual (user-modifiable / operable)" to components (i.e., "important components") whose importance is higher than a threshold (or determined to be important) so that the user can modify and operate them as desired when visualizing the scenario. When the visualization processing unit 1228, described later, visualizes a component labeled "manual (user-modifiable / operable)," it visualizes it in a way that allows user modification and operation. As an example, it is assumed that a 3DCG created in advance is used. Furthermore, the labeling unit 128 assigns an "automatic" label to components whose importance is lower than a threshold (or determined to be unimportant), which is assumed not to be modified or operated by the user. When the visualization processing unit 1228 (described later) visualizes a component labeled "automatic," it automatically generates images and videos from text (attribute information of the component) using, for example, a trained model. As an example, it is expected that attribute information of the location that will be the background of each scene when visualized will be automatically generated, and the characters and props / stage props that appear in the foreground in the video will be visualized using a method (3DCG) that allows easy user modification and manipulation.

[0091] (Component correction part 129) In accordance with user input, the component modification unit 129 performs processing to appropriately modify (update) the information of the component stored in the current component DB 145. For example, from an editing screen for each component, the user can add or modify attribute information of the component, modify the importance, or change the assigned label.

[0092] (Direction proposal department 1226) The performance suggestion unit 1226 proposes the performance content to be visualized based on the extracted / estimated (and further labeled) components, and writes out instructions to the command generation unit 1227 in a text file.

[0093] Specifically, the direction suggestion unit 1226 first determines which components to automatically generate and which components to visualize so that the user can modify and operate them, based on the labels assigned to the components. Next, the direction suggestion unit 1226 uses the information about the components, data from past works, and the importance of each component to propose direction such as sound, lighting, camerawork, etc. for each scene.

[0094] Sounds are obtained from the analysis results of scenario information by the component extraction unit 121B and the component estimation unit 126. Examples of sounds include everyday sounds, environmental sounds, and animal cries. The presence or absence of sound and its intensity are estimated mainly by combining the components of location and props with the verbs or modifiers of those components. For example, the direction suggestion unit 1226 searches for and proposes appropriate sound source files from the general knowledge database 147, etc., because both "the intercom rings" and "the bell rings" produce sound but have different tones. In addition, for the description "heavy rain," the direction suggestion unit 1226 converts the degree of modification into a numerical value and proposes the sound of rain along with the intensity. Lighting is proposed based on location information (attribute information of the component "location") obtained from the analysis results of scenario information, props and stage equipment related to lighting, and other effects (sound and camerawork). When the scenario (script) specifies direct camerawork, such as "OST (over the shoulder)" which refers to an "over the shoulder shot" used when filming characters, that data is expected to be used. On the other hand, if there is no specification in the scenario, the direction suggestion unit 1226 will propose screen composition, angle camera movement, etc., taking into consideration the viewpoint of the characters in each scene and the importance of the elements.

[0095] The proposed production content (text-based) is presented to the user on the display unit by the output control unit 123B, and is modified by the user as appropriate. Proposals for production content are made for each scene, but consideration is also given to references to past works and the consistency of the entire scenario.

[0096] The general knowledge DB 147 stores various knowledge data such as people's physiques and clothing, sizes and colors of tools, speed of movements, sound source files, etc. The general knowledge DB 147 also stores data used in extraction by the component extraction unit 121B and estimation by the component estimation unit 126 (for example, data associating names with nicknames (ANDREA and ANDY, etc.)), which may be referred to as appropriate in the extraction and estimation processes.

[0097] (Command generation unit 1227) The command generation unit 1227 has a function of converting text-based information such as information about the components and instruction contents (performance contents) created by the performance proposal unit 1226 into visualization commands that can be read by the visualization engine (and can be processed by the visualization processing unit 1228). The performance proposal unit 1226 described above writes out the performance contents for each scene based on the components as text in a predetermined data format. Therefore, this content needs to be converted into a predetermined command so that it can be read and visualized by the visualization engine (visualization processing unit 1228). The command DB 148 stores the generated commands (converted data) and supplementary information used during conversion. The supplementary information used during conversion is information added to the command in order to issue more detailed instructions to the visualization processing unit 1228. For example, when the instruction content includes the predicate and object “go to the west” of the action in the time-series metadata (see Figure 30), which is one of the components, and the default coordinates of the subject, the command generation unit 1227 is expected to convert it into a command that adds the content “move” as a type at the time of visualization, or to convert it into a command that embodies abstract expressions or movement coordinates and directions that are not written in the linguistic information.

[0098] (Visualization processing unit 1228) The visualization processing unit 1228 is a visualization engine that generates a simulation image of a scenario, i.e., performs a scenario visualization process. Specifically, the visualization processing unit 1228 reads commands output from the command generation unit 1227 and executes visualization processing. Specifically, the visualization processing unit 1228 searches for 3DCG that can be modified and operated by the user and automatically generates other components, and visualizes the "components for visualization" after gathering them. The modifiable and operable 3DCG may be created and prepared in advance. Furthermore, such 3DCG is assumed to be simple CG for simulation images. In the automatic generation, the visualization processing unit 1228 may input attribute information of the components into a generative model (trained model) and output mainly 2D / 3D. Furthermore, in the case of a movie scenario, the visualization processing unit 1228 visualizes the searched 3DCG and automatically generated images for each scene. The user can modify and operate the appearance and position of the modifiable and operable components (3DCG) using the input unit 11.

[0099] (Search processing unit 1229) The search processing unit 1229 performs processing to search for predetermined information from the past work component DB 146 based on the keyword input by the user. The search results are displayed on the display unit by the output control unit 123B. The search results may also be used to calculate shooting costs, CG production costs, etc. Details will be described later with reference to Figs. 38 to 41.

[0100] (Output control unit 123B) The output control unit 123B controls the display of the information generated by the output information generation unit 122B (for example, a simulation image visualized using simple CG) on a display unit (an example of the output unit 13).

[0101] An example of the configuration of the control unit 12B according to the second embodiment has been described above. Note that the configuration shown in Fig. 21 is just an example, and the present embodiment is not limited to this. For example, it is not necessary to have all of the configuration shown in Fig. 21.

[0102] <4-2. Operation processing> FIG. 22 is a flowchart showing an example of the overall flow of the operation processing according to the second embodiment.

[0103] 22, first, the component extractor 121B acquires data of the scenario (current work) (step S203). For example, the input unit 11 may input the main body of the scenario (text data).

[0104] Next, component extraction unit 121B extracts component data from the scenario data (step S206) and stores the extracted component data in current work component DB 145 (step S209). Here, Figs. 25 to 27 show examples of attribute definitions in the information of each component. Fig. 25 is a diagram showing an example of a character table, Fig. 26 is a diagram showing an example of a location table, and Fig. 27 is a diagram showing an example of a stage / props table. Data of each component is extracted from the scenario data by component extraction unit 121B, and attribute information is filled in (attributes are defined).

[0105] A UUID (Universally Unique Identifier) ​​is an identifier for unique identification. The component extraction unit 121B also performs identical element determination during the extraction process, assigning uniquely identifiable identifiers to components such as characters, locations, and props. The identical element determination can be performed by natural language processing (such as syntax analysis) of scenario information. For example, when the same character has a different name and is extracted as separate components, the UUID can be used to identify (associate) the same character. Attribute information such as the "Name" and "Person ID" in FIG. 25, the "Scene Number" in FIG. 26, and the "Name" and "Props" in FIG. 27 are keys for uniquely managing data during the extraction process and the estimation process. For example, characters and props can be managed by name during extraction and by ID during estimation. Each component shown in FIGS. 25 to 27 corresponds to overall metadata and is ultimately stored in the past work component DB 146. That is, the information on each component element stored in the past work component DB 146 is also formed by tables such as those shown in Figures 25 to 27. Note that a specific example of the data configuration of the past work component DB 146 will be described later with reference to Figure 29.

[0106] Here, the determination of identical elements will be explained with reference to FIG. 28. As shown in the upper part of FIG. 28, for example, in the character table of work A (component "character data"), each character with a different name is determined to be a different element. However, it is determined by analyzing the scenario information (or by modification or addition by the user) that the name "AA-MAN" is the name after transformation of the name "PETER." In this case, a character relationship table such as that shown in the lower part of FIG. 28 is generated and stored in the current work component DB 145. Furthermore, the component extraction unit 121B can determine that the element with the name "PETER" and the element with the name "AA-MAN" are the same person and associate them by assigning the same UUID.

[0107] The above-mentioned components such as characters, locations, and props / stage props are overall metadata that are independent of scenes. The component extraction unit 121B according to this embodiment also extracts scene-dependent time-series metadata as a component. Time-series metadata is an element related to the time series (for example, the content written in "stage directions," which are sentences that instruct actions and performances). FIG. 30 is a diagram showing an example of attribute definitions of time-series metadata (components). FIG. 31 is a diagram showing an example of detailed definitions of "sentence elements" included in the time-series metadata. The component extraction unit 121B extracts time-series metadata such as those shown in FIGS. 30 and 31 from scenario information and stores it in the current work component DB 145. The overall metadata and the time-series metadata are associated with each other using scene numbers and UUIDs.

[0108] Next, if the data of the component has an undefined attribute (step S212 / Yes), the component estimation unit 126 estimates the data (attribute information) of the component (step S215) and stores the estimated data of the component in the current work component DB 145 (step S218). An undefined attribute occurs when the attribute information of the component shown in each table such as those shown in FIGS. 25 to 27 has not been filled in. The component estimation unit 126 can estimate the attribute information of the component by analyzing text information (stage directions and dialogue), referring to the general knowledge DB 147, or referring to the past work component DB 146. Machine learning may also be used for the estimation process.

[0109] FIG. 29 shows an example of the data configuration of the past work component DB 146. As shown in FIG. 29, the past work component DB 146 stores, for each work, a character table, a character relationship table, a location table, a props / stage props table, and the like. The component estimation unit 126 can refer to this past work component DB 146. For example, in the case of a series of works, it is also conceivable that attribute information can be obtained by referring to the past work component DB 146 from the work title and the names of the characters.

[0110] Next, the importance determination unit 127 determines the importance of each component element, and based on the determination result, the label assignment unit 128 assigns an automatic / manual label (step S221). Details of the importance determination and label assignment will be described later with reference to FIG.

[0111] Next, the rendering suggestion unit 1226 determines, based on the label of each component, which components are to be automatically generated and which components are to be modifiable and operable (step S224).

[0112] Next, the rendition suggestion unit 1226 presents rendition details such as sound, lighting, and camera work for each scene to the user (step S227).

[0113] Next, the control unit 12B accepts user modifications to the elements and the content of the rendering from the input unit 11 (step S230). The elements can be modified, for example, from an element editing screen (see FIGS. 35 to 37). As an example, it is possible to change the automatic / manual label.

[0114] Next, the rendition proposing unit 1226 writes out (generates a text file) instructions (elements, rendition details) to be output to the command generating unit 1227 (step S233). If the user modifies the elements, the rendition proposing unit 1226 reflects the modifications in the current work element DB 145. At this point, the control unit 12B may also transfer the overall metadata of the elements stored in the current work element DB 145 to the past work element DB 146.

[0115] Next, the command generation unit 1227 generates a command for visualization (converts the instruction in the text file into a command) (step S236) based on the instruction output from the performance suggestion unit 1226. Specifically, the command generation unit 1227 performs command conversion of the overall metadata and command conversion of the time-series metadata.

[0116] Furthermore, the command generating unit 1227 complements the command as necessary (step S239). Specifically, the command generating unit 1227 complements the command with reference to the command DB 14 in order to output a more detailed instruction to the visualization processing unit 1228.

[0117] Next, the command generating unit 1227 stores the generated command in the command DB 14 (step S242).

[0118] Next, the visualization processing unit 1228 searches for modifiable and operable 3DCG (assets) in response to the command (step S245). Specifically, for components that have been determined to be important and have been manually labeled, a search is performed for modifiable and operable 3DCG (assets).

[0119] Furthermore, the visualization processing unit 1228 automatically generates images using the generative model (trained data) in response to the command (step S248). Specifically, automatic image generation (visualization) is performed for components that are determined to be unimportant and have been automatically labeled.

[0120] Then, the visualization processing unit 1228 visualizes each scene (step S251). The video generated by the visualization is presented to the user on the display unit by the output control unit 123B. Audio may also be output at the same time.

[0121] Furthermore, the control unit 12B receives user corrections to the visualized image from the input unit 11 (step S254). Specifically, the posture, position, etc. of the 3DCG that can be corrected and operated and is included in the image can be corrected.

[0122] As described above, in the scenario creation support according to the second embodiment, when a scenario is visualized, the importance of each component extracted from the scenario is determined, and the components are visualized separately as components that are visualized in a modifiable and operable manner and components that are automatically generated. Automatic generation can be performed, for example, using a generative model, but this does not always result in correct output. In this embodiment, important components are visualized in a modifiable and operable manner, further improving user convenience. Furthermore, it is possible to reflect user modifications as appropriate between text analysis and visualization, thereby supporting better scenario creation.

[0123] (Details of extraction process) Next, the extraction process according to this embodiment will be specifically described with reference to Fig. 23. Fig. 23 is a flowchart showing an example of the flow of the extraction process according to the second embodiment. Here, as an example, extraction of information related to "characters" among the components will be described.

[0124] As shown in FIG. 23, upon acquiring scenario data (step S303), the component extracting unit 121B extracts stage directions, dialogue, and the like from the scenario data (step S306).

[0125] Next, the element extracting unit 121B extracts a list of the names of the characters from the stage directions (step S309).

[0126] Next, the component extraction unit 121B determines whether or not gender can be determined from the name (step S312). For example, it may refer to a gender determination dictionary stored in the general knowledge DB 147. The gender determination dictionary is dictionary data in which words that can determine gender and their genders are stored in pairs, such as "WOMAN:Female, MAN:Male, he:Male, she:Female."

[0127] Next, if the gender can be determined (step S312 / Yes), the component extracting section 121B stores the corresponding gender as component data in the current component DB 145 (step S315).

[0128] On the other hand, if the gender cannot be determined (step S318 / No), the component extracting unit 121B performs anaphora analysis of the stage directions (step S318).

[0129] If the component extraction unit 121B can determine the gender from the anaphor by referring to the gender determination dictionary (step S321 / Yes), it stores the corresponding gender as component data in the current component DB 145 (step S324).

[0130] On the other hand, if gender cannot be determined from the anaphor (step S321 / No), the component extraction unit 121B performs syntactic analysis of the stage directions to extract appositive words (step S327). For example, the component extraction unit 121B associates "MAY" with "aunt" in the sentence "MAY is Peter's aunt."

[0131] If the component extraction unit 121B can determine the gender from the co-positional words by referring to the gender determination dictionary (step S330 / Yes), it stores the corresponding gender as component data in the current component DB 145 (step S333).

[0132] On the other hand, if gender cannot be determined from the apposition words (step S330 / No), the component extraction unit 121B stores "gender: undefined" as component data in the current component DB 145 (step S336).

[0133] (Details of estimation process) Next, the estimation process according to this embodiment will be specifically described with reference to Fig. 24. Fig. 24 is a flowchart showing an example of the flow of the estimation process according to the second embodiment. Here, as an example, estimation of the gender of a "character" among the components will be described.

[0134] As shown in FIG. 24, first, the element estimation unit 126 acquires a character table from the current work element DB 145 (step S353).

[0135] Next, the element estimation unit 126 identifies the character IDs of the characters whose gender is "undefined" (step S356). If there is no data whose gender is "undefined," this process ends.

[0136] Next, the element estimation unit 126 acquires the extracted attribute information of the corresponding person ID from the character table (step S359), and determines whether there is attribute information related to gender (step S412).

[0137] Next, if there is attribute information related to gender (step S412 / Yes), the component estimation unit 126 estimates the gender using a classification model from the related attribute information (step S415).

[0138] On the other hand, if there is no attribute information related to gender (step S412 / No), the component estimation unit 126 acquires dialogue parts extracted from the scenario data (step S418). The headings, stage directions, dialogue parts, etc. extracted from the scenario data by the component extraction unit 121B may be stored in the current work component DB 145.

[0139] Next, the component estimation unit 126 identifies and estimates lines from which gender can be estimated (step S421). For example, the output "Male" is obtained from the line "I'm a busy man, Mr. Parker."

[0140] If gender data can be output by estimation using the classification model or estimation from the dialogue (step S424 / Yes), the component estimation unit 126 stores the corresponding gender as component data in the current work component DB 145 (step S427). On the other hand, if gender data cannot be output (step S424 / No), this process ends.

[0141] (Details of importance determination and label assignment) Next, the importance determination and label assignment process according to this embodiment will be described. The importance of each component can be calculated based on, for example, the number of times the component appears. By determining the importance of each component, visualization can be divided into automatically generated components and components that can be modified and manipulated. Components that require particular attention to detail (such as props) can be listed and presented to the user. The user can view important components and identify those that are likely to be captured as frames using camerawork. Calculating the importance of each component for each scene, in addition to the overall importance, can provide the user with information to consider areas to be removed when visualizing the scene (for example, removing a scene in the middle where only important components appear overall is redundant). The worldview of the components modified and manipulated (including additions and creations) by the user can also be reflected in the automatically generated components for the same scene.

[0142] The definition of "important" can be, for example, (1) if it appears frequently (an element that appears in many scenes; Frequent Element), (2) if it is a key element to the entire story (an element with detailed depictions; Crucial Element), or (3) if it is a key element in a scene (a Focus Element that has a close relationship with a component of the Crucial Element (e.g., the protagonist) in that scene). The importance of (1) can be determined based on the number of times it appears in the entire scenario (the entire story). The importance of (2) can be determined based on whether it is described in detail, i.e., whether there is a lot of extracted attribute information. The importance of (3) is determined for each scene. Specifically, for example, in the case of a "character," it can be determined based on whether the character has many lines in the scene, whether there are many stage directions related to the character, etc. In addition, in the case of a "location," it can be determined based on whether the character or props / stage props are unimportant (e.g., do not appear) in the scene. In addition, in the case of a "prop / stage prop," it can be determined based on whether it has a close relationship with the Crucial Element (e.g., a character), for example, whether it is touched by the Crucial Element.

[0143] The flow of the operational process will be described below with reference to Fig. 32. Fig. 32 is a flowchart showing an example of the flow of the importance determination and label assignment process according to the second embodiment.

[0144] As shown in FIG. 32, first, the importance determination unit 127 acquires data of the components extracted from the scenario (from the component extraction unit 121B or current work component DB 145) (step S433).

[0145] Next, the importance determination unit 127 determines the overall (entire scenario) importance of each component (step S436). The above-mentioned "Frequent Element" and "Crucial Element" are assumed to be definitions of overall importance. For example, for a "Frequent Element," if the number of appearances exceeds a threshold, the importance determination unit 127 determines the corresponding component as important (or of high importance, "Main"). For a "Crucial Element," if the number of pieces of extracted attribute information exceeds a threshold (e.g., 10 or more), the importance determination unit 127 determines the corresponding component as important (or of high importance, "Main"). The importance determination unit 127 may also calculate the importance of each component by taking into account the perspectives of "Frequent Element" and "Crucial Element." For example, the importance determination unit 127 may calculate the importance (number of points) of each component by weighting the number of scenes in which each component appears according to the proportion of the extracted attributes of that component. If the importance level is equal to or greater than a predetermined number of points, the component is determined to be an important component.

[0146] Next, the importance determination unit 127 updates the current component DB 145 so as to add the determined importance as attribute information of the component (step S439).

[0147] The importance determination unit 127 can accept a user's modification of the importance from the input unit 11 (step S442).

[0148] Next, the importance determination unit 127 performs processing for each scene based on the scenario information. Specifically, first, data on the components appearing in the target scene is obtained (step S445).

[0149] Next, the importance determination unit 127 analyzes the header, stage directions, and dialogue of the scenario in the target scene (step S448).

[0150] Next, the importance determination unit 127 determines the importance of the constituent element in the target scene. The aforementioned "Focus Element" is assumed as a definition of the importance of a scene. Regarding the "Focus Element," an element that is closely related to the "Crucial Element," for example, may be considered important (a predetermined number of points indicating the importance may be given), or the number of attribute information extracted from the scene may be greater than a threshold. Furthermore, a constituent element that uses a predetermined feature term using machine learning may be considered important. Furthermore, the importance determination unit 127 may calculate the importance of the constituent element in a scene by taking into account the overall importance of the constituent element that has already been calculated. For example, the importance determination unit 127 may determine the final importance of the constituent element as the sum of the overall importance and the importance of the scene calculated using the above method. Furthermore, the importance determination unit 127 may also add the importance of the previous scene. If the importance of the constituent element is equal to or greater than a predetermined number of points, the importance determination unit 127 determines the constituent element to be an important element. Furthermore, the importance determination unit 127 may rank all the components appearing in a scene according to their importance, or may rank them by category.

[0151] Next, based on the result of the importance determination, if the component can be determined to be an important element (for example, the importance point exceeds a threshold value) (step S454 / Yes), the label assignment unit 128 assigns a label (manual label) that indicates that the component can be modified or operated (step S460).

[0152] On the other hand, if it is not an important element (step S454 / No), the labeling unit 128 assigns an automatically generated label (automatic label) (step S457).

[0153] Then, the label assignment unit 128 updates the current component DB 145 so as to add the assigned label as attribute information of the component (step S463).

[0154] The labeling unit 128 can accept a change (change) of a label by the user from the input unit 11 (step S446). The user can manually change the label between automatic and manual for each component on the editing screen (see FIGS. 33, 35 to 37).

[0155] The above has described the importance determination and label assignment processes according to this embodiment. Note that the importance determination and label assignment processes according to this embodiment are not limited to this. For example, the user may manually assign an automatic / manual label to each extracted component element on an editing screen (see FIGS. 33, 35 to 37).

[0156] Furthermore, when a scenario is modified, the importance determination unit 127 may re-determine the importance of the modified scene and update the current work component DB 145. Furthermore, the importance determination unit 127 may periodically analyze the entire scenario and update the overall importance and the importance of each scene.

[0157] <4-3. Display screen example> FIG. 33 is a diagram showing an example of an editing screen that allows editing the importance and labels of components in a scene according to the second embodiment. As shown in screen 600 in FIG. 33, when a cursor is placed over a word in the main scenario text, if the component has been extracted and its importance has been determined, a display 601 indicating its importance (whether it is an important element or not) is displayed. The extracted components are listed by category on the right side of the screen. The left side of the screen displays a display 603 of important elements in the scene. A ranking of important elements by category and a ranking of all important elements are displayed here. The user may correct the importance by rearranging the important element rankings, deleting a component from the important element rankings, or dragging a component from the main scenario text into the Focus Elements area on the screen. The main scenario text can also be corrected on screen 600. When an update button 605 on the screen is selected, the importance determination unit 127 reanalyzes the scene and updates the importance of the component.

[0158] Fig. 34 is a diagram showing an example of a confirmation screen for the importance of each scene in which a component appears according to the second embodiment. Screen 610 in Fig. 34 displays a graph showing the importance of the scene in which "music box" appears, and a table showing the importance ranking for each scene. In the graph showing the importance of a scene, it is also possible to select multiple components and display them simultaneously. By referring to the scenes in which the components appear in the entire scenario and the importance of each scene in which they appear, the user can estimate the filming costs (filming time) and schedule.

[0159] 35 to 37 are diagrams showing examples of editing screens for each component. Specifically, screen 620 shown in FIG. 35 is an editing screen for a stage prop (door), which is an example of a component. Extracted attribute information 622 has already been entered here, and the user can manually correct it as needed or enter information in fields that have not been extracted. A 2D image 621 automatically selected based on the attribute information of the component may be displayed for reference. The user may upload any 2D image 621 or delete it. Selecting a "Calculate" button 623 may also perform an estimate of the CG production cost of the component. The "CG production cost of the component" may be an estimate of the production cost of the CG used in the actual footage, or an estimate of the production cost of the CG used in the simulation footage. Details of the CG production cost estimate will be described later with reference to FIG. 40.

[0160] In addition, the components can be displayed in order of frequency in the lower left corner of the screen. When the user has finished inputting the attribute information for each component, he / she selects the “Create” button to reflect the input contents in the current component DB 145.

[0161] Also, on the left side of the screen 620, a display 625 is displayed showing a label (automatic / manual) assigned to the component indicating the production method for visualization. If the user wants to change the label (automatic / manual), the user selects a "Change to Manual" button 624 displayed in the lower right of the screen 620. Selecting the "Change to Manual" button 624 transitions the screen to an editing screen 620m shown in FIG. 36. At this time, the label of the component has been changed from automatic to manual. If the user wants to return it to automatic, he or she selects a "Change to Automatic" button 626. Also, a "CG Search" button 627 is displayed on the editing screen 620m. When the "CG Search" button 627 is selected, a search is performed for 3DCG that the user can modify and operate. 3DCG that the user can modify and operate may be generated in advance.

[0162] As an optional feature, it is also possible to switch to a performer search screen or a location search screen.

[0163] Figure 37 is an editing screen for a character, which is an example of a component. Extracted attribute information 632 has already been entered on screen 630 shown in Figure 37, and the user can manually correct it as needed or enter information in fields that have not been extracted. Also, a 2D image 631 automatically selected based on the attribute information of the component may be displayed for reference. Furthermore, by selecting a "Calculate" button 633, the CG production costs for the component may be estimated.

[0164] In the example shown in FIG. 37, the label "manual" is added as shown in display 635, but if the user wants to change this to automatic generation, he or she selects "change to automatic" button 634.

[0165] Furthermore, by selecting the "Search for Performer" button 636, a search for an appropriate performer is performed based on the attribute information of the characters, and the search results are displayed. Performer search will be described with reference to FIG.

[0166] Although an example of the editing screen has been described above, the screen configurations shown in FIGS. 35 to 37 are merely examples, and the present embodiment is not limited to these.

[0167] <4-4. Application Examples> Next, a description will be given of an application example of the second embodiment. In this embodiment, the performance of various searches and trial calculations can be improved based on attribute information of each component element extracted / estimated from scenario information.

[0168] (Location-based search for past works) FIG. 38 is a flowchart showing an example of the flow of location-based search processing for past works according to an application example of the second embodiment. In the process of producing a movie, there is a need to refer to the settings and images of past works, such as "I want to check the camerawork in a movie that has a car chase scene." When relying only on human memory or using web searches, there is a problem that the desired information cannot be found or takes a long time. In this embodiment, information on each extracted / estimated component element is managed in association with the work name (e.g., "Work name ___ location table"), thereby improving search efficiency.

[0169] As shown in FIG. 38, first, the search processing unit 1229 receives a search term (for example, "Eiffel's Tower") input by the user from the input unit 11 (steps S503 and S505).

[0170] Next, the search processing unit 1229 acquires the location table of all works from the past work component DB 146 (step S509).

[0171] Next, if there is a location table containing a component with the location name "Eiffel's Tower" (step S512 / Yes), the search processing unit 1229 identifies the location table name (e.g., "Work A_Location Table") (step S515), and further identifies the work name (e.g., "Work A") (step S518).

[0172] Then, the search processing unit 1229 displays a list of the identified work titles as the search results (step S521). If there is no location table containing an element whose place name is "Eiffel's Tower" (step S512 / No), the search processing unit 1229 displays "No Matches" as the search results.

[0173] (Performer search processing) 39 is a flowchart showing an example of the flow of processing for actor search according to an application example of the second embodiment. Conventionally, when searching for actors, it was necessary for a person to read the scenario and determine the necessary conditions for the actor (gender, age, height, etc.), but in this embodiment, by using the attribute information of the characters extracted / estimated from the scenario to search actor databases inside and outside the system, it is possible to reduce the time required to check the scenario and the effort required for manual search.

[0174] As shown in FIG. 39, first, the search processing unit 1229 receives input from the user of the work name, characters (element elements: characters), and role class ("Main") (step S533).

[0175] Next, the search processing unit 1229 acquires the character table of the relevant work from the past work component DB 146 (step S536). Note that data on the components of the scenario for which component extraction / estimation and visualization processing have been performed has been transferred to the past work component DB 146 from the current work component DB 145. Even information about a work that has not yet been actually filmed can be stored as information about an analyzed work.

[0176] Next, if there is a person whose role class is "Main" (step S539 / Yes), the search processing unit 1229 identifies the corresponding person ID (step S542) and obtains attribute information of the identified person ID from the character table (step S545).

[0177] Next, the search processing unit 1229 displays the acquired attribute information and presents it to the user (step S548). The search processing unit 1229 also searches for performers corresponding to the acquired attribute information from performer databases inside and outside the system, and displays the search results and presents them to the user (step S551). Performer searches may be performed on an external website.

[0178] If there is no person whose role class is "Main" (step S539 / No), the search processing unit 1229 displays "No Match" as the search result (step S554).

[0179] (CG production cost calculation process) Figure 40 is a flowchart showing an example of the process flow for CG production cost estimation using an application example of the second embodiment. CG is often used in previsualization and in the final video. However, apart from reading the script, preparing to order and estimating costs for CG production requires time and effort. After a human reads the script and finds the objects they want to turn into CG, such as ordering individual body parts for characters, they must then further break them down into smaller elements for ordering CG parts. This system automatically extracts and estimates objects, thereby reducing time. Furthermore, because it extracts and estimates elements in units of visualization components, it is closer to the units required for creating CG parts than typical human perception. This provides information useful for CG ordering and production not only for professional video creators but also for general users. Furthermore, if script analysis results can predict the cost of creating CG for a project, it can reduce the time required for script selection. By comparing the estimated estimate with the budget, it is possible to individually consider objects and scenes to be converted into CG.

[0180] 40, first, the search processing unit 1229 receives input of the work title, characters, and hair length from the user (step S563). In this flow, as an example, a case will be described in which "the cost of CG production of all characters with long hair appearing in a work is estimated."

[0181] Next, the search processing unit 1229 acquires the character table of the relevant work from the past work component DB 146 (step S566).

[0182] Next, if there is a person with "long" hair length (step S569 / Yes), the search processing unit 1229 counts the number of corresponding individuals (step S572) and displays the number of individuals (step S575). The search processing unit 1229 also calculates the cost of CG production using a CG production cost estimation engine based on the attribute information and number of individuals of the target characters, and displays the calculation result (step S578). The CG production cost estimation engine assumes a database and calculation model that calculates the cost and man-hours required to produce 3DCG to visualize each component.

[0183] If there is no person with "long" hair length (step S569 / No), the search processing unit 1229 displays "no match" as the search result (step S581).

[0184] (Filming cost calculation process) FIG. 41 is a flowchart showing an example of the processing flow of the photographing cost trial calculation according to the application example of the second embodiment.

[0185] 41, first, the search processing unit 1229 receives input of the title of the work, the location, and whether or not it exists in reality from the user (step S603). In this flow, as an example, a case where "the name of an existing place is acquired and the shooting cost is estimated" will be described.

[0186] Next, the search processing unit 1229 acquires the location table of the relevant work from the past work component DB 146 (step S606).

[0187] Next, if there is a location for which the existence / non-existence is "True" (step S609 / Yes), the search processing unit 1229 identifies the corresponding location ID (step S612) and acquires the location name and country (step S618). An example of a location database according to this application example is shown in Fig. 42. As shown in Fig. 42, in the location database, the scene number in which the location appears, existence / non-existence, country, city, location name, and location ID are associated as attribute information of the location.

[0188] Next, the search processing unit 1229 displays the acquired attribute information (step S618). The search processing unit 1229 also calculates the filming costs taking into account the filming locations (locations) and displays the calculation results (step S621). The filming costs are calculated using a filming cost estimation engine. The filming cost estimation engine is assumed to perform processing to calculate the costs of filming locations, etc., using a database or an external website that stores filming locations and costs, etc.

[0189] If there is no location for which the existence status is "True" (step S609 / No), the search processing unit 1229 displays "No hits" as the search result (step S624).

[0190] As described above, by estimating the approximate cost of filming based on the location, it is possible to reduce the time required for selecting a scenario and considering a budget. In addition, the attribute information acquired in step S618 (such as the name of the actual location, country, city, and scene number) is displayed, which is useful when creating a location filming schedule.

[0191] In addition, the acquired attribute information also includes the scene number in which the location appears, so for example, by sorting the search results by location name, it is possible to obtain a list of scenes in which the same location appears. This reduces the time required for tasks such as coordinating location shooting dates and shooting scenes. The above search process can also be used in cases where schedule management is software-based.

[0192] (Box office forecast) The output information generation unit 122B according to this embodiment can predict box office revenue when new scenario information is input by using learning data of past works including audience evaluation scores and expert evaluation scores. By referring to the box office revenue prediction when selecting a scenario, the selection time can be reduced. Furthermore, when writing a scenario, the box office revenue prediction can be checked as one indicator while writing. Furthermore, when producing a movie based on a novel, the box office revenue prediction can be referred to during the production process.

[0193] It is assumed that box office revenue predictions are calculated by creating a box office prediction model using a supervised learning method that uses a neural network. Furthermore, learning data for building the model includes information about the characters and the setting, as well as audience evaluation scores and expert evaluation scores for previously released works. In this embodiment, a portion of the learning data is obtained by performing an analysis, such as extracting / estimating information about components, on a large number of previous works. Then, a model is built based on this data, along with various evaluation data for previous works, and even when a new scenario is input, a box office revenue prediction can be presented to the user as one of the analysis results.

[0194] <<5. Supplementary Information>> Although the preferred embodiments of the present disclosure have been described in detail above with reference to the accompanying drawings, the present technology is not limited to such examples. It is clear that a person skilled in the art of the present disclosure can conceive of various modified or altered examples within the scope of the technical ideas described in the claims, and it is understood that these also naturally fall within the technical scope of the present disclosure.

[0195] For example, one or more computer programs can be created to cause hardware such as a CPU, ROM, and RAM built into the information processing device 1 to perform the functions of the information processing device 1. Also provided is a computer-readable storage medium storing the one or more computer programs.

[0196] Furthermore, the effects described herein are merely descriptive or exemplary and are not limiting. In other words, the technology according to the present disclosure may achieve other effects that will be apparent to those skilled in the art from the description of this specification, in addition to or in place of the above-described effects.

[0197] The present technology can also be configured as follows. (1) a process of estimating a value generated by the content based on information of one or more configuration elements configured to generate the content; A process of comparing the estimated value with a target value; a process of outputting correction information regarding correction of the setting element based on the result of the comparison; An information processing device comprising a control unit that performs the above. (2) The information processing device according to (1), wherein the correction information is information regarding an increase or decrease in the number of the setting elements. (3) The information processing device according to (2), wherein the control unit determines a setting element to be added or deleted depending on a difference between the estimated value and the target value. (4) The information processing device according to (3), wherein the control unit determines the setting element to be added or deleted using learning data of content generated in the past. (5) The information processing device according to (4), wherein the control unit determines a setting element to be deleted from the one or more setting elements. (6) The information processing device according to (4), wherein the control unit determines a setting element to be added from one or more setting elements of the content generated in the past. (7) The information processing device according to any one of (1) to (6), wherein the information on the setting elements is information on characters in a story, character relationships, locations, eras, props, or stage props. (8) The information processing device according to (7), wherein the control unit performs a process of extracting information about the setting elements from information about a scenario. (9) The information processing device according to (8), wherein the value generated by the content is a time length of a video. (10) The information processing device according to (8), wherein the value generated by the content is a shooting cost or a profit. (11) The information processing device according to any one of (1) to (10), wherein the control unit generates one or more pieces of correction information and controls a display unit to display the generated one or more pieces of correction information as modification proposals. (12) The information processing device according to (11), wherein the control unit performs control to display one or more pieces of correction information as a card-type UI. (13) The setting elements are extracted from scenario information; The information processing device described in any one of (1) to (12), wherein the control unit, when the correction information is adopted by a user, controls to display the range of scenarios that will be affected by the execution of the correction information. (14) The control unit determining the importance of a component element corresponding to the setting element, which is used when generating a simulation video based on information about the scenario, based on information about the scenario; The information processing device according to (1), wherein a component to be visualized and operable by a user is determined from among the one or more components according to the importance. (15) The processor: estimating a value to be generated by the content based on information of one or more configuration elements configured to generate the content; comparing the estimated value with a target value; outputting correction information regarding correction of the setting element based on a result of the comparison; An information processing method, including: (16) Computer, a process of estimating a value generated by the content based on information of one or more configuration elements configured to generate the content; A process of comparing the estimated value with a target value; a process of outputting correction information regarding correction of the setting element based on the result of the comparison; A program that functions as a control unit to perform the above. [Explanation of symbols]

[0198] 1. Information processing equipment 11 Input section 12 (12A, 12B) Control section 121 Element Extraction Unit 122 (122A, 122B) Output information generation section 123 (123A, 123B) Output control section 13 Output section 14 Storage section 121A Setting element extraction unit 124 Tagging Processing Unit 1221 Estimation Department 1222 Comparison section 1223 Correction information generation unit 1224 Display screen generation section 141 Past works knowledge DB 142 Past work setting element DB 143 Configuration element change history DB 121B Component extraction part 126 Component Estimation Unit 127 Important Judgment Section 128 Labeling section 1226 Direction Proposal Department 1227 Command Generation Unit 1228 Visualization Processing Unit 145 Current work component DB 146 Past work component DB 147 General knowledge DB 148 Command DB148

Claims

1. a process of estimating a value generated by the content based on information of one or more setting elements configured to generate the content; A process of comparing the estimated value with a target value; a process of outputting correction information regarding correction of the setting element based on the result of the comparison; An information processing device comprising a control unit that performs the above.

2. The information processing apparatus according to claim 1 , wherein the correction information is information relating to an increase or decrease in the number of the setting elements.

3. The information processing apparatus according to claim 2 , wherein the control unit determines a setting element to be added or deleted depending on a difference between the estimated value and the target value.

4. The information processing device according to claim 3 , wherein the control unit determines the setting element to be added or deleted by using learning data of content generated in the past.

5. The information processing device according to claim 4 , wherein the control unit determines a setting element to be deleted from the one or more setting elements.

6. The information processing device according to claim 4 , wherein the control unit determines the setting element to be added from one or more setting elements of the previously generated content.

7. The information processing device according to claim 1 , wherein the information on the setting elements is information on characters in a story, character relationships, locations, periods, props, or stage props.

8. The information processing device according to claim 7 , wherein the control unit performs a process of extracting information about the setting elements from information about a scenario.

9. The information processing device according to claim 8 , wherein the value generated by the content is a time length of a video.

10. The information processing device according to claim 8 , wherein the value generated by the content is a photography cost or a profit.

11. The information processing apparatus according to claim 1 , wherein the control unit generates one or more pieces of correction information and controls a display unit to display the generated one or more pieces of correction information as a change proposal.

12. The information processing device according to claim 11 , wherein the control unit performs control to display one or more pieces of correction information as a card-type UI.

13. The setting elements are extracted from scenario information; The information processing device according to claim 1 , wherein the control unit performs control to display a range of a scenario that will be affected by execution of the correction information when the correction information is adopted by a user.

14. The control unit determining the importance of a component element corresponding to the setting element, which is used when generating a simulation video based on information about the scenario, based on information about the scenario; The information processing apparatus according to claim 1 , further comprising: determining, from among the one or more components, a component to be visualized and operable by a user, according to the importance level.

15. The processor: estimating a value to be generated by the content based on information of one or more configuration elements configured to generate the content; comparing the estimated value with a target value; outputting correction information regarding correction of the setting element based on a result of the comparison; An information processing method, including:

16. Computer, a process of estimating a value generated by the content based on information of one or more setting elements configured to generate the content; A process of comparing the estimated value with a target value; a process of outputting correction information regarding correction of the setting element based on the result of the comparison; A program that functions as a control unit to perform the above.

Images