System for managing music content based on ai

Abstract

The AI ​​sound source management system of the present invention includes an AI sound source management device comprising: a user terminal that connects to an AI sound source management platform and provides sound source conditions input by a user to the AI ​​sound source management platform; and a processor that operates the AI ​​sound source management platform and provides sound source content corresponding to the sound source conditions through a pre-trained artificial intelligence model. The processor includes an AI composition unit that generates sound source content based on composition conditions included in the sound source conditions; an AI score arrangement unit that arranges an original score based on arrangement conditions included in the sound source conditions to generate an AI arranged score; and a virtual ensemble practice unit that receives individual audio files in which a user’s practice is recorded from the user terminal and generates a virtual ensemble piece using individual audio files of valid users.

Description

Technology Field

[0001] The present invention relates to an AI sound source management system. Background Technology

[0002] Church worship is a communal liturgical act offered to God, and praise serves as a core element that shapes the atmosphere of worship and expresses emotions of faith. Beyond being mere musical devices, hymns play a role in helping worshippers focus their hearts and, along with the sermon, emotionally reinforce the message of the service. In particular, selecting appropriate hymns according to the season or sermon theme has a significant impact on the overall atmosphere and flow of the service, and through this, the choir or praise team performs the important ministry of leading the congregation's confession of faith through music.

[0003] The production and use of hymns in worship settings largely rely on manual work and experience, which is causing various structural problems. In particular, due to a lack of sufficient audio content aligned with sermon themes or seasons (e.g., Lent, Pentecost, Advent), the reality is that similar repertoires of hymns are repeatedly used year after year.

[0004] This problem is also caused by the structure of music content supply; the church music market is dominated by a few large publishers and composers, and the limited musical styles and arrangement methods make it difficult to make flexible choices tailored to various age groups and worship atmospheres.

[0005] Furthermore, since choirs are generally composed of members of diverse ages and skill levels, some members may experience difficulties with reading sheet music or maintaining a sense of rhythm. It is difficult for the conductor to monitor individual practice progress in real time and to provide repeated feedback on areas needing improvement. Consequently, the actual quality of the ensemble performance inevitably suffers relative to the practice time, posing a problem that directly impacts the musical quality of the worship service. The problem to be solved

[0006] The technical objective of the present invention is to provide an AI sound source management system capable of resolving the inefficiencies and content limitations of the existing worship music environment by automating the processes of sound source content creation, practice, arrangement, recommendation, and distribution. means of solving the problem

[0007] An AI sound source management system according to an embodiment of the present invention comprises an AI sound source management device including a user terminal that connects to an AI sound source management platform and provides sound source conditions input by a user to the AI ​​sound source management platform, and a processor that operates the AI ​​sound source management platform and provides sound source content corresponding to the sound source conditions through a pre-trained artificial intelligence model. The processor comprises an AI composition unit that generates sound source content based on composition conditions included in the sound source conditions, an AI score arrangement unit that arranges an original score based on arrangement conditions included in the sound source conditions to generate an AI arranged score, and a virtual ensemble practice unit that receives individual audio files in which a user’s practice is recorded from the user terminal and generates a virtual ensemble piece using individual audio files of valid users.

[0008] According to an embodiment, the AI ​​composition unit may compose a chord progression and a melody using at least one of season information, theme information, worship message summary, and mood information included in the composition conditions, and may determine the rhythmic complexity, the range of melody leaps, whether ornaments are used, and the number of modulations using at least one of conductor style, performer composition information, and choir member level included in the composition conditions.

[0009] According to an embodiment, the AI ​​composition unit analyzes a reference sound source included in the composition conditions to determine the style of a conductor who conducted the performance of the reference sound source, generates sound source content that reflects rhythm interpretation, rubato processing methods, and musical notation usage habits similar to the determined conductor style, and the conductor style may include at least one of the conductor's rhythm interpretation, rubato processing methods, and musical notation usage habits.

[0010] According to an embodiment, the AI ​​composition unit quantifies the melody line, harmonic progression, rhythm pattern, instrument arrangement, song development structure, tonality, tempo, and dynamics of the generated sound source content to generate embedding information, and determines similarity by comparing the embedding information with the embedding information of existing sound source content stored in a database using any one of cosine similarity, dynamic time warping (DTW), and triplet loss-based similarity determination networks, and if the generated sound source content has high similarity to the existing sound source content, the generation of sound source content can be repeated until a preset similarity passing criterion is satisfied.

[0011] According to an embodiment, the AI ​​score arrangement unit can perform vocal part adjustment, vocal range rearrangement, harmonic structure supplementation, and instrument arrangement by considering at least one of the actual vocal range of the choir included in the arrangement conditions, instrument configuration, performance proficiency of the members, and conductor style.

[0012] According to an embodiment, the AI ​​score arrangement unit analyzes a reference sound source included in the arrangement conditions to determine the style of the conductor who conducted the performance of the reference sound source, generates the AI ​​arranged score reflecting rhythm interpretation, rubato processing methods, and musical notation usage habits similar to the determined conductor style, and provides a display that allows visual confirmation of changes caused by the arrangement by comparing the AI ​​arranged score with the original score.

[0013] According to an embodiment, the virtual ensemble practice unit may generate the virtual ensemble piece by applying time axis alignment and tempo correction to each individual audio file of the valid users, and performing volume balancing, spatial reverb processing, and noise removal.

[0014] According to an embodiment, the virtual ensemble practice unit can accumulate feedback of the virtual ensemble piece generated by the conductor in a database to generate statistical data including individual proficiency changes, number of practice sessions, and accuracy improvement rates.

[0015] According to an embodiment, the AI ​​sound source management system further includes a sound source recommendation unit that recommends similar sound source content in response to recommendation conditions included in the sound source conditions, and when a reference sound source content is included in the recommendation conditions, the sound source recommendation unit analyzes the musical structure, tonality, beat, rhythm pattern, harmonic progression, MR sound type, and sentiment analysis of the reference sound source content, calculates a similarity score between the analysis result and existing sound source content stored in a database, and recommends the similar sound source content by sorting them in order of high similarity score.

[0016] According to an embodiment, the AI ​​sound source management system may further include a transaction management unit that supports transactions and management regarding the distribution, copyright management, revenue settlement, permission for secondary use, and donation linkage of the sound source content on the AI ​​sound source management platform. Effects of the invention

[0017] According to the AI ​​sound source management system of the embodiment of the present invention, the processes of creating, practicing, arranging, recommending, and distributing sound source content specialized for church worship can be automated, thereby resolving the inefficiency and content limitation issues of the existing worship music environment.

[0018] In addition, according to the AI ​​sound source management system of the embodiment of the present invention, customized sound source content can be automatically generated based on composition conditions entered by the user, allowing praise songs optimized for a specific time or message to be produced and used immediately.

[0019] In addition, according to the AI ​​sound source management system of the embodiment of the present invention, practice sound sources of individual units are collected and automatically synthesized into a single ensemble sound source, thereby enabling remote ensemble practice.

[0020] In addition, according to the AI ​​sound source management system of the embodiment of the present invention, sheet music can be automatically arranged according to arrangement conditions entered by the user, so that sheet music suitable for the choir's vocal composition, instrument holdings, and member performance levels can be easily obtained. Brief explanation of the drawing

[0021] FIG. 1 is a drawing of an AI sound source management system according to an embodiment of the present invention. FIG. 2 is a schematic block diagram of an AI sound source management device according to an embodiment of the present invention. FIG. 3 is a schematic block diagram of a processor according to an embodiment of the present invention. FIG. 4 is a flowchart illustrating an AI sound source management method according to an embodiment of the present invention. FIG. 5 is a flowchart illustrating an AI sound source management method according to another embodiment of the present invention. Specific details for implementing the invention

[0022] The advantages and features of the present invention and the methods for achieving them will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below but can be implemented in various different forms. These embodiments are provided merely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the present invention, and the present invention is defined only by the scope of the claims.

[0023] The terms used in this specification are for describing embodiments and are not intended to limit the invention. In this specification, the singular form includes the plural form unless specifically stated otherwise in the text. The terms "comprises" and / or "comprising" as used in this specification do not exclude the presence or addition of one or more other components in addition to the components mentioned. Throughout the specification, the same reference numerals refer to the same components, and "and / or" includes each of the mentioned components and all combinations of one or more. Although terms such as "first," "second," etc., are used to describe various components, these components are not limited by these terms. These terms are used merely to distinguish one component from another. Therefore, the first component mentioned below may be the second component within the technical scope of the invention.

[0024] Unless otherwise defined, all terms used herein (including technical and scientific terms) may be used in a meaning commonly understood by those skilled in the art to which the present invention pertains. Additionally, terms defined in commonly used dictionaries are not to be interpreted ideally or excessively unless explicitly and specifically defined otherwise.

[0025] The terms “part” or “module” as used in the specification refer to software or hardware components, such as FPGAs or ASICs, and “part” or “module” perform certain roles. However, “part” or “module” is not limited to software or hardware. “Part” or “module” may be configured to reside in an addressable storage medium or configured to run on one or more processors. Thus, by example, “part” or “module” includes components such as software components, object-oriented software components, class components, and task components, as well as processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables. The functions provided within the components and “parts” or “modules” may be combined into a smaller number of components and “parts” or “modules,” or further separated into additional components and “parts” or “modules.”

[0026] Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings.

[0027] FIG. 1 is a drawing of an AI sound source management system according to an embodiment of the present invention.

[0028] Referring to FIG. 1, an AI sound source management system (10) according to an embodiment of the present invention automates the entire process of composition, arrangement, practice, recommendation, distribution, and trading occurring in music activities such as worship, choir, and choir based on artificial intelligence, and provides an AI sound source management platform service that generates and manages sound source content optimized for the user's music environment in real time.

[0029] An AI sound source management system (10) according to an embodiment of the present invention provides various sound source content-related services to a user through an AI sound source management platform and includes an AI sound source management device (100) and a user terminal (200).

[0030] The AI ​​sound source management device (100) is a device capable of hosting an online network and network addressing, and may be an operating server that provides an AI sound source management service online. The AI ​​sound source management device (100) may communicate with a user terminal (200) through a network communication network to provide an AI sound source management service.

[0031] Here, the network communication refers to a connection structure capable of exchanging information between the AI ​​sound source management device (100) and the user terminal (200), and may include a local area network (LAN), a wide area network (WAN), the internet (WWW), a wired and wireless data communication network, a telephone network, a wired and wireless television communication network, etc.

[0032] The AI ​​sound source management device (100) may be a cloud computing model that operates an AI sound source management platform, builds an AI sound source management platform on-premise, and provides a Platform as a Service (PaaS) capable of running and managing related applications.

[0033] Additionally, the AI ​​sound source management device (100) can also be implemented as Software as a Service (SaaS) that allows users to immediately use the function through a web or mobile application without separate installation.

[0034] The AI ​​sound source management device (100) comprehensively controls all processes ranging from the creation, practice, recommendation, distribution, and trading of sound source content carried out on an AI sound source management platform, and can interpret user input conditions using a pre-trained artificial intelligence model and provide sound source content corresponding to the conditions. Here, the sound source conditions may include composition conditions for creating sound source content, arrangement conditions for creating AI arrangement scores, and recommendation conditions for recommending similar sound source content.

[0035] Artificial intelligence models can be deep learning models. A deep learning model (e.g., a deep neural network (DNN)) may refer to a multilayer perceptron that includes multiple hidden layers in addition to the input and output layers. Deep neural networks can be used to identify the latent structures of data. Deep neural networks may include, but are not limited to, convolutional neural networks (CNN), recurrent neural networks (RNN), autoencoders, Generative Adversarial Networks (GAN), restricted Boltzmann machines (RBM), deep belief networks (DBN), Q networks, U networks, and Siamese networks.

[0036] The AI ​​sound source management device (100) can automatically generate sheet music and MR (Music Resource) optimized for the user and provide the results to the user through an AI sound source management platform. In particular, the AI ​​sound source management device (100) can generate sound content by considering various factors such as worship seasons, sermon topics, choir composition, conductor style, and performance difficulty.

[0037] The AI ​​sound source management device (100) can collect individual audio files in which the members' practice is recorded, generate a virtual ensemble piece that is close to an actual ensemble performance, and generate an analysis report based on the virtual ensemble piece, thereby providing quantitative feedback data to the conductor. At this time, the AI ​​sound source management device (100) automatically sorts, corrects, and analyzes the individual audio files and merges them into a single virtual ensemble piece, and the conductor can receive the virtual ensemble piece and the individual audio files to directly check pronunciation, pitch, rhythm, etc., or refer to the feedback data provided by the AI ​​to deliver individual feedback to each member.

[0038] The AI ​​audio management device (100) can automatically recommend audio content that matches the flow of worship or the content of the sermon. For example, the AI ​​audio management device (100) can learn the user's past worship information, praise song selection history, preferred genre, instrument configuration, etc., and select and present to the user songs that are similar in emotional context or musical characteristics.

[0039] The AI ​​sound source management device (100) automatically handles the distribution, usage tracking, copyright protection, and revenue settlement of sound source content registered on the AI ​​sound source management platform. When sheet music or sound source registered by a user is used by another church or user, the usage history of the sound source content is automatically saved, and revenue can be distributed according to the contribution. Individual user vocal sound sources and instrument performance data can also be subject to distribution immediately upon registration, and the AI ​​sound source management device (100) can perform automatic settlement based on usage frequency or part utilization ratio.

[0040] The user terminal (200) is a terminal for various users such as conductors, members, worship leaders, and content consumers, and can access the AI ​​sound source management platform through a web browser or a dedicated application and perform the creation, practice, download, and trading of sound source content on the AI ​​sound source management platform.

[0041] According to an embodiment, the user terminal (200) may mean a PC, a smartphone, a tablet PC, a mobile internet device (MID), an internet tablet, an IoT (internet of things) device, an IoE (internet of everything) device, a desktop computer, a laptop computer, a workstation computer, or a PDA (personal digital assistant), but is not limited thereto.

[0042] Through the user terminal (200), the user can request the creation of sound source content by inputting various composition conditions, such as worship seasons, sermon topics, instrument configurations, vocal distribution of the choir, and performance difficulty, and the requested composition conditions can be provided to the AI ​​sound source management device (100). The sound source content created by the AI ​​sound source management device (100) can be downloaded from the user terminal (200), and the user terminal (200) can provide various user convenience functions, such as automatic scrolling of the sheet music, viewing by part, and switching to practice mode.

[0043] In particular, the user terminal (200) can support the practice environment of each unit. For example, a unit member can select their own vocal part or instrument part and practice while playing a practice MR optimized for that part, and the practiced content can be recorded and saved as an audio file or transmitted to an AI sound source management platform.

[0044] The conductor can check the audio file through their user terminal (200), provide feedback to individual members, or guide them while listening to a virtual ensemble piece.

[0045] FIG. 2 is a schematic block diagram of an AI sound source management device according to an embodiment of the present invention.

[0046] Referring to FIG. 2, an AI sound source management device (100) according to an embodiment of the present invention includes a processor (110), memory (120), a communication interface (130), and storage (140).

[0047] The processor (110) controls the overall operation of each component of the AI ​​sound source management device (100). The processor (110) may be configured to include a CPU (Central Processing Unit), an MPU (Micro Processor Unit), an MCU (Micro Controller Unit), a GPU (Graphic Processing Unit), or any form of processor (110) well known in the art of the present invention.

[0048] Additionally, the processor (110) can perform operations for at least one application or program for executing the method according to embodiments of the present invention, and the server (100) may have one or more processors (110).

[0049] According to an embodiment, the processor (110) may further include RAM (Random Access Memory, not shown) and ROM (Read-Only Memory, not shown) for temporarily and / or permanently storing signals (or data) processed within the processor (110). Additionally, the processor (110) may be implemented in the form of a system-on-chip (SoC) comprising at least one of a graphics processing unit, RAM, and ROM.

[0050] Memory (120) stores various data, commands, or information. Memory (120) may load a program from storage (140) to execute a method according to various embodiments of the present invention. When a computer program is loaded into memory (120), the processor (110) may perform the method by executing one or more instructions that constitute the computer program. Memory (120) may be implemented as volatile memory such as RAM, but the technical scope of the present disclosure is not limited thereto.

[0051] The communication interface (130) supports wired or wireless communication of the AI ​​sound source management device (100). Additionally, the communication interface (130) may support various communication methods other than internet communication. To this end, the communication interface (130) may be configured to include a communication module well known in the art of the present invention. In some embodiments, the communication interface (130) may be omitted.

[0052] Storage (140) can store computer programs non-temporarily. When providing AI sound source management services through the AI ​​sound source management device (100), storage (140) can store various information required for generation and processing during the execution of the process.

[0053] The storage (140) may be configured to include non-volatile memory such as ROM (Read Only Memory), EPROM (Erasable Programmable ROM), EEPROM (Electrically Erasable Programmable ROM), flash memory, a hard disk, a removable disk, or any form of computer-readable recording medium well known in the art to which the present invention belongs.

[0054] The bus (150) provides communication functions between components of the AI ​​sound source management device (100). The bus (150) can be implemented as various types of buses, such as an address bus, a data bus, and a control bus.

[0055] FIG. 3 is a schematic block diagram of a processor according to an embodiment of the present invention.

[0056] Referring to FIG. 3, the processor (110) includes an AI composition unit (111), an AI score arrangement unit (112), a virtual ensemble practice unit (113), a sound source recommendation unit (114), and a transaction management unit (115), each component being implemented by a software module, a hardware module, or a combination thereof, and may mean a unit that is connected through at least one communication path and processes at least one function or operation.

[0057] The AI ​​composition unit (111) can automatically generate sound content optimized for worship situations based on composition conditions included in the sound source conditions through a pre-trained artificial intelligence model. Beyond simply automatically creating music, the AI ​​composition unit (111) can generate sheet music and MR together at a level that can be immediately used in actual worship settings by comprehensively reflecting specific situations and contexts required by the church, characteristics of the members, the conductor's style, and usage environment. Here, the sound content may include lyrics, sheet music, and MR.

[0058] The AI ​​composition department (111) can compose a chord progression and melody using at least one of the season information, theme information, worship message summary, and mood information included in the composition conditions to convey musical identity and emotion suitable for the context of worship, and can determine the complexity of the rhythm, the range of melody leaps, whether to use ornaments, and the number of modulations using at least one of the conductor style, the composition information of the performers, and the level of the choir members included in the composition conditions.

[0059] For example, composition conditions may include thematic information and mood information such as ‘Advent’, ‘atmosphere of waiting and hope’, ‘youth group worship’, ‘female vocal part centered’, reference audio of past conductors to reflect the conductor’s style, and composition of members such as soprano, alto, and tenor, and the AI ​​composition unit (111) can perform composition that optimizes the vocal range and instrument arrangement in response to the composition conditions.

[0060] Specifically, the AI ​​composition unit (111) can receive seasonal information corresponding to the worship situation from the user and can generate sound content based on the seasonal information. Here, the seasonal information can be subdivided into Christmas, Easter, Lent, Advent, Harvest Festival, etc., and each season can contain a unique theological message and emotional atmosphere.

[0061] The AI ​​composition unit (111) can generate sound content based on topic information input by the user. Here, the topic information may be input in the form of text written directly by the user or provided as topic keywords selectable by the user. The AI ​​composition unit (111) can use the topic information to design the composition of lyrics and melody lines, harmonic progression methods, and musical narrative structure.

[0062] According to an embodiment, the AI ​​composition unit (111) may receive a summary of a worship message from a user, generate a draft of lyrics based on the summary of the worship message, and generate audio content based on the structure of the lyrics. For example, if a summary of a worship message with the theme "God is with us" is received, the AI ​​composition unit (111) may automatically generate a draft of praise lyrics based on the theme "God is with us," and compose a chord progression and melody suitable for the praise to complete the entire audio content.

[0063] The AI ​​composition unit (111) can generate sound content based on mood information input by the user. Mood information is a condition for precisely adjusting the emotional expression of the sound source and may include a majestic mood, a reverent mood, a joyful mood, a sad mood, a victorious mood, a peaceful mood, etc.

[0064] The AI ​​composition unit (111) can generate sound content that reflects the tempo, beat, chord tone, and melody progression method based on mood information. For example, if the user selects a majestic mood, the AI ​​composition unit (111) can generate sound content that has a wide interval between notes and a slow and strong tempo rhythm, with a structure that starts at a low pitch range and gradually rises.

[0065] The AI ​​composition unit (111) can generate audio content that reflects the conductor's style. The AI ​​composition unit (111) can receive reference audio from the user terminal (200) and can determine the conductor's style, such as performance speed, dynamic processing, and dynamic control patterns, based on the reference audio. For example, the AI ​​composition unit (111) can receive reference audio from the user terminal (200), such as choral recording files or MIDI data that the conductor has previously directed. By analyzing the reference audio, it can generate audio content that reflects the conductor's rhythm interpretation, rubato processing method, and habits of using musical notation. Through this, the AI ​​composition unit (111) can generate audio content containing a unique interpretation that reflects the characteristics of the conductor of each church.

[0066] The AI ​​composition unit (111) can generate sound content by reflecting the composition information of the performer. Here, the composition information of the performer may include the vocal parts possessed by the choir, participating instruments, gender, age group, etc.

[0067] For example, if the choir using the sound source is a female choir centered on soprano and alto and possesses only piano and violin, the AI ​​composition unit (111) can compose the piece in an SSA (soprano-soprano-alto) or piano structure, excluding male parts for tenor and bass, and arrange the melody within a limited range suitable for the female parts. Additionally, the AI ​​composition unit (111) can generate sound source content composed mainly of violin melody and piano accompaniment, omitting the sounds of brass instruments or low-pitched string instruments.

[0068] For example, if an adult choir possesses all four SATB parts but has few members in the tenor and bass parts and the instrument playing level is low, the user can set the tenor and bass parts as supporting parts or simplified parts. In this case, the AI ​​composition unit (111) can generate sound content by focusing the main melody and harmony on the soprano and alto parts, and giving only repeating rhythms or simple harmonic tones to the tenor and bass parts, thereby maintaining the balance of the entire song while reducing the burden on the actual performers.

[0069] The AI ​​composition unit (111) can generate sound sources by taking into account the level of the unit members. The user can separately specify the average performance level of the entire unit, the performance level of a specific instrument, the choral proficiency of the entire vocal part, and the choral proficiency of a specific vocal part, etc., in grades of beginner, intermediate, and advanced. Based on the unit level input from the user terminal (200), the AI ​​composition unit (111) can generate sound source content that takes into account the complexity of the rhythm, the range of melody leaps, whether ornaments are used, the number of modulations, etc.

[0070] For example, if the level of the entire unit is set to beginner, the AI ​​composition unit (111) can simplify the rhythm of the entire score to mainly quarter notes and eighth notes, not use fast tempos or complex time signatures (e.g., 5 / 4, 7 / 8), and limit leap intervals within a voice part to within a major third. In addition, the AI ​​composition unit (111) can exclude ornaments (such as trills, turns, and melismas including slurs) and generate sound content with a syllable-centered structure in which the lyrics and pitch match.

[0071] According to an embodiment, the user may specify the composition format, such as the total length of the song, whether there is an introduction, whether the chorus structure is repeated, whether there is a mixture of solo and chorus, and the language of the lyrics, and the AI ​​composition unit (111) may generate sound content according to the input composition format.

[0072] The AI ​​composition unit (111) can generate sound content by combining the composition conditions entered by the user and utilizing a previously trained artificial intelligence model.

[0073] According to an embodiment, the MR of the sound source content may also include the sound of instruments that the choir does not possess. For example, if a choir that has only string parts needs the sound of brass instruments, the AI ​​composition unit (111) can generate an MR that includes the sound of brass instruments by reflecting a pre-learned sound source library and a conductor style.

[0074] Additionally, the AI ​​composition unit (111) can generate a similarity review result by comparing the similarity between the generated sound content and existing sound content stored in the database. The AI ​​composition unit (111) can perform a style comparison to determine in what respects the generated sound content is similar to or different from existing hymns, chants, classical worship music, etc.

[0075] Specifically, the AI ​​composition unit (111) can vectorize structural, emotional, and technical elements of music and analyze them in a high-dimensional space to determine similarity. In particular, the AI ​​composition unit (111) is based on a music similarity judgment algorithm and a style embedding network, and can prevent composition copyright issues or review whether a song is suitable for a specific worship atmosphere.

[0076] According to an embodiment, the AI ​​composition unit (111) can analyze the melody line, chord progression, rhythm pattern, instrument arrangement, song development structure (verse-chorus-bridge, etc.), key, tempo, dynamics, etc. of the generated sound source content and quantify or pattern each element.

[0077] Melodies can be analyzed using N-gram analysis based on a sequence of pitches or embedded into high-dimensional vectors using a deep learning-based melody encoder, and chord progressions can be converted into degree-based sequences and used for similarity comparison.

[0078] The embedding information can be compared with existing sound content, such as hymns, choral songs, and classical worship music, stored in a database. The AI ​​composition unit (111) can use comparison algorithms such as cosine similarity, dynamic time warping (DTW), and triplet loss-based similarity determination networks. For example, if the cosine similarity between the melody embedding of the generated sound content and the embedding of the existing sound content is 0.92 or higher, the AI ​​composition unit (111) can determine that there is a high degree of similarity to the existing song.

[0079] If the AI ​​composition unit (111) has a high degree of similarity to existing sound content stored in a database, it can repeat the generation of sound content until it satisfies a pre-set similarity passing criterion.

[0080] The AI ​​composition unit (111) can analyze the tonality, rhythm structure, harmonic progression, melody line, mood tag, etc. of the generated sound content to automatically search for similar music in the database and provide a visual comparison result.

[0081] The AI ​​score arrangement unit (112) can generate an AI arranged score by modifying the original score registered in the AI ​​sound source management platform to suit the user's environment and requirements. The AI ​​score arrangement unit (112) can not only perform simple tempo adjustments or key changes, but also comprehensively consider the choir's actual vocal range, instrument composition, member's performance proficiency, and conductor's style to perform adjustments by vocal part, vocal range rearrangement, harmony structure supplementation, and instrument arrangement changes.

[0082] To this end, the AI ​​score arrangement unit (112) can analyze arrangement conditions included in the sound source conditions. The arrangement conditions may include the average and maximum range of each voice part (S, A, T, B), the presence or absence of a specific voice part (e.g., absence of tenor), a list of instruments owned (e.g., presence of cello and piano, absence of brass), preference for playing difficulty (beginner to advanced range), conductor style, etc.

[0083] For example, if the user inputs information that the choir they belong to is a women's choir centered on sopranos and altos, has no tenors or basses, and possesses cellos and violins, the AI ​​can rearrange the existing SATB 4-part score into an SSA (Soprano-Soprano-Alto) configuration and improve the structure by transferring the harmony lines previously handled by the bass or tenor to the cello or low piano parts.

[0084] The AI ​​score arrangement unit (112) can perform range optimization so that each vocal part sings in the most stable range. For example, if the score structure is such that the soprano part repeats excessively high notes or the alto part continuously uses a low range close to a baritone, the AI ​​score arrangement unit (112) can remove unstable ranges or automatically adjust to a more comfortable and resonant structure through octave shifting or harmony rearrangement.

[0085] The AI ​​sheet music arrangement unit (112) can generate an AI arrangement sheet music by emphasizing or weakening specific vocal parts. For example, if the user's arrangement conditions include a request to "emphasize a soft atmosphere centered on the soprano vocal part in this week's worship service," the AI ​​sheet music arrangement unit (112) can generate an AI arrangement sheet music by concentrating the main melody on the soprano, arranging the remaining vocal parts as simple harmony or sound support roles, and placing the soprano line relatively large in the MR. Conversely, if the user's arrangement conditions include a request to "arrange with a grand atmosphere centered on the bass part since the bass part is strong," the AI ​​arrangement sheet music can be generated by dispersing the melody to the lower vocal parts and increasing the proportion of low-pitched instruments.

[0086] Additionally, the AI ​​score arrangement unit (112) can generate an AI arrangement score that can be practiced individually by utilizing the conductor style included in the arrangement conditions. The conductor style is a concept beyond simple tempo values ​​or dynamic notations and can include the overall musical interpretation, such as phrasing (breathing units), rubato processing methods, methods of connecting measures, emphasis sections, and habits of using musical notation, which are frequently used by actual conductors. That is, the AI ​​score arrangement unit (112) can analyze the reference audio source included in the arrangement conditions to determine the conductor style that conducted the performance of the reference audio source, and generate an AI arrangement score that reflects rhythm interpretation, rubato processing methods, and habits of using musical notation similar to the determined conductor style.

[0087] According to an embodiment, the AI ​​score arrangement unit (112) can determine the conductor style by analyzing reference sound sources, such as existing choral sound sources and MIDI files, uploaded to the AI ​​sound source management platform through the user terminal (200).

[0088] For example, if the conductor has a habit of temporarily slowing down the tempo during the chorus of a song, the AI ​​score arrangement unit (112) can generate an AI arranged score with rubato or rit. symbols inserted in that section.

[0089] For example, if the conductor has a style of frequently changing the crescendo and diminuendo in each phrase, the AI ​​score arrangement section (112) can place musical notation symbols in the corresponding measures and reflect the dynamic processing in the MR as well.

[0090] The AI ​​score arrangement unit (112) provides a visual indicator that allows the changes resulting from the arrangement to be checked by comparing the generated AI arranged score with the original score, and may also provide a UI / UX that partially maintains a part of the original song or selectively reflects only specific arrangement elements according to the user's request. Accordingly, the conductor can operate the AI ​​arranged score more flexibly and select the version most suitable for the performance situation.

[0091] The virtual ensemble practice section (113) connects the boundaries between individual practice and the whole ensemble using digital technology, generates a single virtual ensemble piece based on the practice results of each member, and enables the conductor to provide real-time feedback even remotely.

[0092] The virtual ensemble practice section (113) allows the user to learn and adapt in advance to rhythmic discrepancies, balance between parts, and differences in interpretation by the conductor that may occur during actual ensemble playing through AI-arranged sheet music.

[0093] AI arrangement scores reflecting the conductor's style can be provided to each member's user terminal (200), and members can view the AI ​​arrangement scores on their user terminal (200) and practice individually while listening to the MR reflecting the conductor's style. At this time, visual markers (e.g., tempo change section indicator, rubato highlight box, breath note notification, etc.) are applied to the user terminal (200), so that the user can clearly recognize the conductor's style, reduce confusion regarding the flow of practice, and accurately execute the instructed musical emotional expression.

[0094] The user can upload individual audio files in which practice is recorded to an AI sound source management platform via a user terminal (200). At this time, the virtual ensemble practice unit (113) can generate a completed virtual ensemble song by applying time axis alignment, tempo correction, etc., to individual audio files provided by a valid user. Here, a valid user refers to a user pre-configured for the generation of the virtual ensemble song.

[0095] According to an embodiment, the virtual ensemble practice unit (113) can generate a virtual ensemble song by applying time axis alignment and tempo correction to each individual audio file of valid users, and performing volume balancing, spatial reverb processing, noise removal, etc.

[0096] The conductor can receive virtual ensemble pieces and individual audio files from the AI ​​sound source management platform through the user terminal (200), and can identify the rhythm accuracy, pitch deviation, and dynamic processing of individual members by listening to the entire sound or separating and checking the performance of each member. The conductor can provide feedback directly to the members regarding the problematic sections or generate a feedback report from the AI ​​sound source management platform to provide practice guidance.

[0097] That is, the virtual ensemble practice unit (113) generates an AI arrangement score reflecting the conductor's style on each member's user terminal (200), and each member can perform individual practice using the AI ​​arrangement score and upload individual audio files of the recorded practice to an AI sound source management platform. The virtual ensemble practice unit (113) synthesizes a single completed virtual ensemble piece using the individual audio files provided by the members, and the conductor can generate feedback for each member using the virtual ensemble piece and the individual audio files.

[0098] In addition, the virtual ensemble practice unit (113) can accumulate feedback generated by the conductor in a database to generate statistical data such as individual proficiency changes, number of practice sessions, and accuracy improvement rates. This virtual ensemble practice unit (113) overcomes the limitations of an offline environment and provides a collaborative music training environment in which the conductor's style is reflected in the individual members' practice in real time, and individual practice leads to a consistent result for the entire ensemble.

[0099] The sound recommendation unit (114) can search for and recommend sound content corresponding to recommendation conditions included in the sound conditions of the user terminal (200). Beyond simple song selection automation functions, the sound recommendation unit (114) can suggest sound by comprehensively considering the user's sound requests, past usage history, worship themes, seasons, vocal arrangements, performance level, congregational reaction, etc.

[0100] When requesting a sound source, the user can directly set recommendation conditions corresponding to the context of the worship service. For example, the user can input recommendation conditions through the user terminal (200), such as the season (Easter, Advent, Lent, etc.), sermon topic (e.g., restoration, patience, work of the Holy Spirit), type of worship (Sunday morning worship, Wednesday prayer meeting, youth group worship, etc.), church size (small choir, large choir, women's choir, etc.), vocal composition (SATB, SSA, monophonic, etc.), instrument ownership status (keyboard only, string instruments included, MR only usable), average performance level of members (beginner to advanced), and preferred atmosphere (reverence, joy, solemnity, etc.).

[0101] The sound recommendation unit (114) can select sound sources that meet recommendation conditions from among the sound sources stored in the database through recommendation keyword matching.

[0102] According to an embodiment, if the standard music content is included in the recommendation conditions, the sound recommendation unit (114) can analyze the standard music content and recommend similar music content to the user. The sound recommendation unit (114) can analyze the musical structure, tonality, beat, rhythm pattern, harmonic progression, MR sound type, and sentiment analysis results of the standard music content, and calculate a similarity score between the analysis results and existing music content stored in the database. The sound recommendation unit (114) can recommend similar music content to the user by sorting it in order of high similarity score.

[0103] The audio recommendation section (114) can provide additional information such as sheet music preview, MR preview, reason for recommendation, structure information, difficulty of use, reviews and evaluation scores from previous churches, denominational suitability, and examples of similar sermons, along with recommendations for similar audio content.

[0104] The sound source recommendation unit (114) can store this as user-customized recommendation information when a user repeatedly selects a specific song or prefers a certain structure (e.g., major key-centered, 3 / 4 time signature song with a certain tempo), and can recommend a sound source based on the user-customized recommendation information when a request for a sound source is made in the future. For example, if the first user has repeatedly used 'slow tempo traditional hymn-style songs,' a song of a similar format can be automatically prioritized in the next recommendation as well.

[0105] The music recommendation section (114) can provide the user with a response to the music content by analyzing the content selection flow of the entire church based on the user's selection data, providing a real-time ranking of popular songs, ranking the top 10 popular praise songs of the month, and ranking the most selected recovery theme hymns of the year.

[0106] The transaction management department (115) provides transaction and management functions, such as distribution of music content, copyright management, revenue settlement, permission for secondary use, and donation linkage, to implement a trading ecosystem for music content through an AI music management platform.

[0107] The transaction management department (115) can support the registration of audio content created or arranged through the AI ​​audio management platform, as well as audio content created by the user, into the AI ​​audio management platform. During the registration process, the user can input the title of the audio content, song description, sermon topic, seasonal relevance, usage conditions (e.g., whether it is for worship only / available for practice), whether it is paid / free, copyright owner information, and co-creator contribution ratio, and the input information can be stored in the form of metadata.

[0108] When sound content and metadata are registered, the transaction management department (115) can generate a unique identifier and perform tamper-proofing and copyright protection through blockchain or unique hash-based authentication. Sound content registered on the AI ​​sound management platform can be exposed so that other users can explore it, and on the user terminal (200), after checking sheet music previews, MR listening, usage examples, creator introductions, reviews, etc., download, streaming, purchase, or license services of the sound content can be provided.

[0109] For paid content, users can select MyCredit, subscription points, credit cards, simple payment, etc., as a payment method, and after the transaction is completed, the usage history can be recorded on the AI ​​music management platform.

[0110] The transaction management unit (115) can automatically distribute revenue based on the contribution ratio between the copyright holder of the sound content, co-creators (e.g., lyricist, arranger, MR mixer, etc.), and the registrant. For example, if a user inputs lyrics and remasters the MR on the AI ​​sound management platform, it can be set so that 40% of the total revenue is distributed to the AI ​​sound management platform, 30% to the user, and 30% to the MR creator. Through automatic revenue distribution, the transaction management unit (115) enables automatic settlement without manual work, reduces revenue conflicts among creators, and realizes a fair distribution system.

[0111] According to an embodiment, the transaction management unit (115) can track the use of derivative works and remixes to distribute revenue. When a user creates their own arrangement or remix content using specific sound source content, the transaction management unit (115) automatically tracks the original creator's contribution and can redistribute a certain amount of revenue to the original creator whenever the content is traded. For example, if a first user registers a female choir song on an AI sound source management platform, and a second user remixes and uses it as a male choir song, and other users reuse the song, the first user, who is the original creator, can continuously receive revenue.

[0112] The transaction management department (115) can track the use of derivative works and remixes using AI-based creation history and metadata tracking technology.

[0113] Additionally, the transaction management unit (115) can support smart contract-based music content transactions to ensure transaction stability and reliability. Smart contracts are technology that can be configured to automatically execute payment, settlement, distribution to contributors, license registration, and usage restriction when transaction conditions are met, thereby enabling transparent transactions without unnecessary administrative procedures. By storing transaction and usage history in a blockchain or zero-knowledge proof-based data structure through smart contracts, the transaction management unit (115) can prevent transaction forgery or unfair disputes at the source and increase the overall reliability of the platform.

[0114] FIG. 4 is a flowchart illustrating an AI sound source management method according to an embodiment of the present invention.

[0115] Referring to FIG. 4, the AI ​​sound source management device (100) can receive sound source conditions including composition conditions from a user terminal (200) (S100). The composition conditions may include keywords describing the user's request, and may include worship seasons (Advent, Easter, Christmas, etc.), sermon topics (repentance, thanksgiving, comfort, etc.), atmosphere (reverent atmosphere, march style, meditative type, etc.), performance difficulty (beginner, intermediate, advanced), vocal composition (SATB), instrumentation (violin, cello, flute, etc.), conductor style (recording file of an existing conductor's piece), etc.

[0116] And, the AI ​​sound source management device (100) can analyze composition conditions through a pre-trained artificial intelligence model (S110). At this time, the AI ​​sound source management device (100) recognizes keywords entered by the user through Natural Language Processing (NLP) and can quantify or tag composition conditions so that the artificial intelligence model can understand them.

[0117] And, the AI ​​sound source management device (100) can generate sound source content in response to analyzed composition conditions (S120). The sound source content may include dynamic processing or tempo changes that reflect the conductor's style included in the composition conditions, and instrument parts that the choir does not possess may also be filled with virtual instruments.

[0118] In addition, the AI ​​sound source management device (100) can generate a similarity review result by comparing the similarity between the sound source content and the existing sound source content stored in the database (S130). In particular, the AI ​​sound source management device (100) can vectorize the structural, emotional, and technical elements of the sound source content and analyze them in a high-dimensional space to determine similarity, and is based on a music similarity judgment algorithm and a style embedding network, and can prevent composition copyright issues or review whether the song is suitable for a specific worship atmosphere.

[0119] And, if the similarity between the sound content and the existing sound content stored in the database is less than the pre-set similarity passing standard, the AI ​​sound management device (100) can provide it to the user terminal (200) through the AI ​​sound management platform (S140). If the similarity between the sound content and the existing sound content stored in the database is greater than or equal to the similarity passing standard, the AI ​​sound management device (100) can repeat the generation of sound content until the pre-set similarity passing standard is satisfied.

[0120] FIG. 5 is a flowchart illustrating an AI sound source management method according to another embodiment of the present invention.

[0121] Referring to FIG. 5, the AI ​​sound source management device (100) can receive sound source conditions including arrangement conditions from a user terminal (200) (S200). The arrangement conditions may include the average and maximum range of each vocal part (S, A, T, B), the presence or absence of a specific vocal part (e.g., absence of tenor), a list of instruments owned (e.g., presence of cello and piano, absence of brass), preference for performance difficulty (beginner to advanced range), conductor style, etc.

[0122] Additionally, the AI ​​sound source management device (100) can analyze arrangement conditions through a pre-trained artificial intelligence model (S210). The AI ​​sound source management device (100) performs an analysis of the original score based on the arrangement conditions and can determine the overall structure of the original score (section composition, development method), melody line, harmony structure, rhythm pattern, vocal range distribution, and division of roles among vocal parts. Additionally, the AI ​​sound source management device (100) can identify parts of the original score that require arrangement based on the arrangement conditions.

[0123] And, the AI ​​sound source management device (100) can generate an AI arrangement score by modifying the melody line of the original score according to the analyzed arrangement conditions, reconfiguring the arrangement of the vocal parts, lowering the difficulty of playing compared to the existing one, or reducing or emphasizing the role of a specific instrument, and adjusting various musical elements (S220).

[0124] And, the AI ​​sound source management device (100) can provide an AI arrangement sheet music to a user terminal (200) through an AI sound source management platform (S230).

[0125] Although embodiments of the present invention have been described above with reference to the attached drawings, those skilled in the art will understand that the present invention may be implemented in other specific forms without altering its technical concept or essential features. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. Explanation of the symbols

[0126] 10: AI Music Management System 100: AI Music Management Device 110: Processor 111: AI Composition Department 112: AI Score Arrangement Department 113: Virtual Ensemble Practice Club 114: Music Recommendation Section 115: Transaction Management Department 120: Memory 130: Communication Interface 140: Storage 150: Bus 200: User terminal

Claims

Claim 1 The AI ​​sound source management device includes: a user terminal that connects to an AI sound source management platform and provides sound source conditions input by a user to the AI ​​sound source management platform; and a processor that operates the AI ​​sound source management platform and provides sound source content corresponding to the sound source conditions through a pre-trained artificial intelligence model, wherein the processor comprises: an AI composition unit that analyzes a reference sound source of composition conditions included in the sound source conditions to determine the style of a conductor who conducted the performance of the reference sound source, and generates the sound source content by reflecting rhythm interpretation, rubato processing methods, and musical notation usage habits similar to the determined conductor style; and an AI score arrangement unit that arranges an original score based on arrangement conditions included in the sound source conditions to generate an AI arranged score. The AI ​​sound source management system includes a virtual ensemble practice unit that receives individual audio files in which a user’s practice is recorded from the user terminal, applies time axis alignment and tempo correction to each individual audio file of valid users, and performs volume balancing, spatial reverb processing, and noise removal to generate a virtual ensemble piece, and the AI ​​composition unit quantifies the melody line, harmonic progression, rhythm pattern, instrument arrangement, song development structure, tonality, tempo, and dynamics of the sound source content to generate embedding information, determines similarity by comparing the embedding information with the embedding information of existing sound source content stored in a database using any one of cosine similarity, dynamic time warping (DTW), and triplet loss-based similarity determination networks, and repeats the generation of the sound source content until a preset similarity passing criterion is satisfied when the generated sound source content has high similarity to the existing sound source content. Claim 2 In claim 1, the AI ​​composition unit composes a chord progression and a melody using at least one of season information, theme information, worship message summary, and mood information included in the composition conditions, and determines the rhythmic complexity, melody leap range, whether ornaments are used, and the number of modulations using at least one of conductor style, performer composition information, and choir member level included in the composition conditions. Claim 3 delete Claim 4 delete Claim 5 In claim 1, the AI ​​score arrangement unit is an AI sound source management system that performs vocal part adjustment, vocal range rearrangement, harmonic structure supplementation, and instrument arrangement by considering at least one of the actual vocal range of the choir included in the arrangement conditions, instrument configuration, performance proficiency of the members, and conductor style. Claim 6 In claim 1, the AI ​​score arrangement unit analyzes a reference sound source included in the arrangement conditions to determine the style of a conductor who conducted the performance of the reference sound source, generates an AI arranged score that reflects rhythm interpretation, rubato processing methods, and musical notation usage habits similar to the determined conductor style, and provides an AI sound source management system that provides a visual indication of changes caused by arrangement by comparing the AI ​​arranged score with the original score. Claim 7 delete Claim 8 In claim 1, the virtual ensemble practice unit is an AI sound source management system that accumulates feedback of the virtual ensemble piece generated by the conductor in a database to generate statistical data including individual proficiency changes, number of practice sessions, and accuracy improvement rates. Claim 9 In claim 1, the processor further includes a sound source recommendation unit that recommends similar sound source content in response to recommendation conditions included in the sound source conditions, and the sound source recommendation unit analyzes the musical structure, tonality, beat, rhythm pattern, harmonic progression, and MR sound type of the reference sound source content, respectively, when a reference sound source content is included in the recommendation conditions, calculates a similarity score between the analysis result and existing sound source content stored in a database, and recommends the similar sound source content by sorting them in order of high similarity score. Claim 10 In claim 1, the processor further comprises a transaction management unit that manages transactions regarding the distribution, copyright, revenue settlement, and donation linkage of the sound content on the AI ​​sound management platform.

Images