An electronic piano dynamic music score generation and key highlighting guide method and system

Abstract

The application discloses a kind of electronic organ dynamic music score generation and keyboard highlight guiding method and system, method includes receiving user input music and difficulty parameter, and pre-processing;Adopt the multi-audio track separation and chord inference algorithm based on deep learning, extract the music features including melody, chord and rhythm in pre-processed music;Adopt multi-modal Transformer model, generate dynamic electronic organ music score according to the music features and difficulty parameter, and adjust music score content according to note density and rhythm complexity;According to electronic organ music score, the keyboard that needs to be played is indicated using light system on electronic organ;Through the collection user's keying speed and accuracy, evaluate user skill level, and adjust electronic organ music score difficulty according to evaluation result in real time;Adopt music generation Transformer technology to generate accompaniment note sequence, and provide a variety of accompaniment instrument options and style adjustment options according to user preference.The application can realize full-process automation and personalized learning.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of artificial intelligence, and specifically relates to an electronic piano dynamic score generation and key highlighting guidance method and system. BACKGROUND

[0002] With the popularization of intelligent electronic pianos in the field of music education, users' requirements for its functions and learning experience are continuously increasing. However, the existing technology has many shortcomings, which limits its teaching effect. First, the light guidance of the intelligent electronic piano only supports pre-set music, users cannot choose new music to learn, and the music update is slow, which is difficult to meet the diversified needs, and the light function is limited to prompting the key position, lacking rhythm and dynamics guidance. Second, the difficulty of the music is fixed and cannot be adjusted according to the user's skills, which is difficult to adapt to beginners and advanced users. In addition, the system lacks personalized arrangement, and the user's interest and creative demand is not supported. The learning feedback is insufficient, and the user is difficult to correct errors or track progress in real time. In summary, the existing technology has defects in dynamic score generation, difficulty adaptation, personalization and feedback, and needs to be improved. SUMMARY

[0003] The present application provides an electronic piano dynamic score generation and key highlighting guidance method and system, which aims to use artificial intelligence model for music analysis, adaptive difficulty grading, and combine with real-time light guidance technology, to provide intelligent music teaching and personalized learning experience for electronic piano users.

[0004] According to a first aspect of the embodiments of the present disclosure, an electronic piano dynamic score generation and key highlighting guidance method is provided, the method comprising the following steps:

[0005] Receiving user input music and difficulty parameters and preprocessing;

[0006] Using a deep learning-based multi-track separation and chord inference algorithm to extract music features including melody, chord and rhythm from the preprocessed music;

[0007] Using a multi-modal Transformer model to generate dynamic electronic piano scores according to the music features and difficulty parameters, and adjusting the score content according to the note density and rhythm complexity;

[0008] According to the electronic piano score, using the light system on the electronic piano to indicate the keys that need to be played;

[0009] By collecting the user's key speed and accuracy to evaluate the user's skill level, and adjusting the difficulty of the electronic piano score in real time according to the evaluation result;

[0010] The music generation Transformer technology is used to generate accompaniment note sequences, and various accompaniment instrument options and style adjustment options are provided according to user preferences.

[0011] In some embodiments, the method further comprises recording user performance audio, comparing with standard music scores, evaluating pitch and rhythm, and generating improvement suggestions.

[0012] In some embodiments, the difficulty parameter includes note density and rhythm complexity, and the qualified fragments are extracted from the parsed music according to the thresholds of note density and rhythm complexity.

[0013] In some embodiments, the music features including melody, chord and rhythm are extracted from the preprocessed music, including the following steps:

[0014] The Music Transformer based on the Transformer pre-training model is used to process audio time series data using self-attention mechanism;

[0015] The Melody RNN model is used to convert audio to MIDI format to extract main melody note sequences;

[0016] The Chord Recognition CNN model is used to analyze chords in the audio;

[0017] The Onset Detection algorithm is used to detect note starting points and calculate rhythm patterns.

[0018] In some embodiments, a multi-modal Transformer model is used to generate dynamic electronic piano scores according to the music features and difficulty parameters, including the following steps:

[0019] A multi-modal Transformer model is used to fuse audio feature vectors m and difficulty parameter vectors d through vector splicing to generate feature vectors x;

[0020] The Transformer encoder is used to process the fused feature vectors x to generate context representations h;

[0021] The Transformer decoder is used to generate the next note based on the context representation and the previously generated note sequence, and finally form the dynamic electronic piano score.

[0022] In some embodiments, according to the electronic piano score, the light system on the electronic piano is used to indicate the keys that need to be played, specifically including: setting the light brightness proportional to the note intensity; setting the light color wherein , represents the note interval time, The color mapping function is determined.

[0023] In some embodiments, the user's performance audio is recorded, and a convolutional neural network is used to perform audio signal comparison with the standard musical score, evaluate the pitch and rhythm, and generate improvement suggestions, specifically expressed as: pitch evaluation ; rhythm evaluation , represents the pitch of the user's performance, represents the standard pitch, represents the rhythm of the user's performance, represents the standard rhythm, and the comprehensive score wherein , represents the weight, represents the tolerance threshold, and based on the deviation result and the comprehensive score, a personalized suggestion is generated using a rule engine.

[0024] According to a second aspect of the embodiments of the present disclosure, an electronic piano dynamic score generation and key highlighting guide system is provided, which comprises:

[0025] A music input module is configured to receive user input music and difficulty parameters and perform preprocessing.

[0026] A feature extraction module is configured to extract music features including melody, chord, and rhythm from the preprocessed music using a deep learning-based multi-track separation and chord inference algorithm.

[0027] A dynamic score generation module is configured to generate a dynamic electronic piano score based on the music features and difficulty parameters using a multi-modal Transformer model, and adjust the score content based on note density and rhythm complexity.

[0028] A key highlighting guide module is configured to use the light system on the electronic piano to indicate the keys that need to be played based on the electronic piano score.

[0029] A difficulty adaptation module is configured to evaluate the user's skill level by collecting the user's key pressing speed and accuracy, and adjust the electronic piano score difficulty in real time based on the evaluation result.

[0030] A personalized arrangement module is configured to generate accompaniment note sequences using music generation Transformer technology, and provide multiple accompaniment instrument options and style adjustment options based on user preferences.

[0031] According to a third aspect of the embodiments of the present disclosure, an electronic device is provided, which comprises a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to implement the steps of the electronic piano dynamic score generation and key highlighting guide method.

[0032] According to a fourth aspect of the embodiments of the present disclosure, a non-transitory computer-readable storage medium is provided, and the storage medium has stored thereon computer instructions, which, when executed by a processor, implement the steps of the electronic keyboard dynamic score generation and keyboard key highlighting guidance method.

[0033] The electronic keyboard dynamic score generation and keyboard key highlighting guidance method, system, electronic device and storage medium provided by the embodiments of the present disclosure can realize full-process automation and personalized learning, improve the flexibility and interest of music teaching, and are suitable for users with different skills and interests.

[0034] It should be understood that the foregoing general description and the following detailed description are only exemplary and explanatory, and are not limiting to the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0035] The accompanying drawings, which are incorporated into and form part of the specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0036] Figure 1 is a flowchart of the electronic keyboard dynamic score generation and keyboard key highlighting guidance method in the embodiments of the present disclosure;

[0037] Figure 2 is a feature extraction flowchart in the embodiments of the present disclosure;

[0038] Figure 3 is a dynamic electronic keyboard score generation flowchart in the embodiments of the present disclosure;

[0039] Figure 4 is a structure diagram of the electronic keyboard dynamic score generation and keyboard key highlighting guidance system in the embodiments of the present disclosure;

[0040] Figure 5 is a schematic diagram of an electronic device in the embodiments of the present disclosure. DETAILED DESCRIPTION

[0041] The present disclosure will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present disclosure, but not to limit the present disclosure. In addition, it should be noted that, for the convenience of description, only the parts related to the present disclosure are shown in the drawings, rather than all the structures.

[0042] Before any embodiments are described in further detail, it should be noted that some embodiments are described as processes or methods, although other embodiments can also be implemented as or with artificial intelligence or machine learning modules, engines, or systems. Although a flow diagram can describe a process as a sequential process, many of the steps can be performed in parallel, concurrently, or with partial concurrence. In addition, the order of the steps can be re-arranged. A process can correspond in part to a software application, function, routine, subroutine, or the like. The various elements of a process can be used alone, in combination, or in a variety of arrangements not specifically discussed in the examples. Although each step in a process can be performed by a single component of a system, the steps can be performed by multiple components operating in coordination. Although a process can describe a process as a whole, various steps or elements of the process can be used in a different process.

[0043] Embodiments of the present application are directed to an electronic piano dynamic score generation and key highlighting guide method and system, electronic equipment and storage medium, aiming to solve the technical defects existing in the existing electronic piano teaching system and improve the user learning experience. In the prior art, the guide function of the intelligent electronic piano is usually limited to pre-set music, lacks flexibility, and cannot adjust the difficulty or provide personalized arrangement support according to user needs. The present application provides the following embodiments:

[0044] An electronic piano dynamic score generation and key highlighting guide method, as shown in Figure 1 The method comprises the following steps:

[0045] S1, receiving user input music and difficulty parameters and preprocessing;

[0046] Specifically, audio file uploading and microphone recording are supported, and the input music is preprocessed, including noise reduction and standardization, to ensure the accuracy of the analysis. Users can upload music and set the learning difficulty, and the system generates personalized teaching content according to the input.

[0047] In one specific embodiment, users can upload music files (such as MP3, WAV format) or record audio through a microphone, and set the learning difficulty (beginner, intermediate, advanced). According to the difficulty selected by the user, the required score fragments are filtered or generated.

[0048] Difficulty preset: receive difficulty selection through the user interface (UI), and the system filters the score according to the following thresholds:

[0049] Beginner: , ,

[0050] Intermediate: , ,

[0051] Advanced: , ,

[0052] Feature calculation: note density: ; rhythm complexity: ;

[0053] Score screening: According to the difficulty selection of the user, the above difficulty threshold is used to extract the qualified fragments from the parsed music, or generate the corresponding teaching content.

[0054] At the same time, the specific expression of complexity evaluation for music fragments or teaching content is: , wherein: , According to the dynamic adjustment of the weight of the music style. , , , Statistical mean and standard deviation of note density and rhythm complexity.

[0055] S2, using a multi-track separation and chord inference algorithm based on deep learning, extracts music features including melody, chord and rhythm from the preprocessed music;

[0056] Specifically, S2 includes track separation, melody extraction, chord analysis and rhythm analysis, which are respectively used to separate the input music into different tracks, extract the main melody, analyze and detect the chord and rhythm pattern. Provide a basis for subsequent score generation.

[0057] In one specific embodiment, the input music is analyzed by artificial intelligence technology to extract key features such as melody, chord and rhythm, providing a data basis for subsequent processing.

[0058] Model selection: Use Music Transformer (a pre-trained model based on Transformer) to efficiently process audio time series data using self-attention mechanism.

[0059] Feature extraction: Melody extraction: use Melody RNN model to convert audio to MIDI format and extract main melody note sequence; Chord recognition: use Chord Recognition CNN model to analyze chords in audio; Rhythm analysis: detect note starting point through Onset Detection algorithm and calculate rhythm pattern.

[0060] Self-attention mechanism of Transformer model: , wherein Query, key and value matrices, is the dimension of the key.

[0061] S3, using a multi-modal Transformer model, generates dynamic electronic piano scores according to the music features and difficulty parameters, and adjusts the score content according to the note density and rhythm complexity;

[0062] Specifically, it supports generating sheet music in multiple difficulty levels, including simplified, standard, and advanced versions, and adjusts the sheet music content according to note density and rhythmic complexity.

[0063] In one specific embodiment, a dynamic musical score is generated based on musical characteristics and user-defined difficulty, and a key light guidance scheme is designed, with the brightness and color changes of the lights used to indicate rhythm and playing dynamics.

[0064] Model selection: A multimodal Transformer model is adopted, which fuses the audio feature vector m and the difficulty parameter vector d through vector concatenation. The fused input is represented as follows: ,in This represents a vector concatenation operation. It is an audio feature vector (dimension n). It is a difficulty parameter vector (dimensions) ).

[0065] Music score generation: Input: The fused feature vector x. Output: The sequence of musical notes. ,in Indicates the first The pitch and duration of each note.

[0066] The Transformer encoder processes the input x and generates a context representation h: h = TransformerEncoder(x);

[0067] The Transformer decoder is based on h and the preceding note sequence s. t-1 Generate the next note s t :s t =TransformerDecoder(h,s t-1 );

[0068] Light brightness With note dynamics Proportional: ,in It's the dynamics of the note. It is the proportionality coefficient.

[0069] Light color: ,in , The interval between notes. This is a color mapping function. For example, fast tempos are indicated by red, and slow tempos by blue. The unit is “notes per second”.

[0070] S4. According to the electronic keyboard sheet music, use the lighting system on the electronic keyboard to indicate the keys that need to be played;

[0071] Specifically, the position of the key is indicated by light, and the visual prompt of rhythm and dynamics is supported, wherein the light brightness is proportional to the dynamics of the note, and the light color changes according to the rhythm speed. That is, the music score is generated and the user is guided to play through the key light (including brightness change).

[0072] S5, evaluate the user skill level by collecting the user's key pressing speed and accuracy, and adjust the electronic piano music score difficulty in real time according to the evaluation result;

[0073] S6, generate accompaniment note sequences by using music generation Transformer technology, and provide multiple accompaniment instrument options and style adjustment options according to user preferences.

[0074] Specifically, adjust the accompaniment and music style to meet the user's creation and learning needs.

[0075] The purpose of the embodiment of the present application is to break through the limitations of traditional electronic piano teaching through intelligent and personalized design, and to improve the flexibility, interest and efficiency of learning.

[0076] In one specific embodiment, the accompaniment content or music style is adjusted according to user needs, such as adding different instrument accompaniment or realizing style transfer. Specifically, music generation Transformer technology is used to support conditional generation. Based on the input main melody and the style label selected by the user , the accompaniment note sequence is output by generation. For example, the user selects a popular style, and the system adds drum and bass accompaniment to the melody.

[0077] The method further comprises S7, by recording the user's playing audio, comparing with the standard music score, evaluating the pitch and rhythm, and generating improvement suggestions. That is, the user's performance is recorded and targeted suggestions are provided to help the user improve skills.

[0078] Specifically, the difficulty parameter in S1 includes note density and rhythm complexity, and the fragments meeting the conditions are extracted from the parsed music according to the threshold of note density and rhythm complexity.

[0079] Specifically, the music features including melody, chord and rhythm in the preprocessed music in S2 are extracted, such as Figure 2 as shown, including the following steps:

[0080] S21, use Music Transformer based on Transformer pre-training model to process audio time series data by using self-attention mechanism;

[0081] S22, convert the audio into MIDI format using the Melody RNN model to extract the main melody note sequence;

[0082] S23, analyze the chords in the audio using the Chord Recognition CNN model;

[0083] S24, detect the onset point of the note by the Onset Detection algorithm and calculate the rhythm pattern.

[0084] In S3, a multi-modal Transformer model is used to generate a dynamic electronic piano score according to the music features and difficulty parameters, as shown in Figure 3 , including the following steps:

[0085] S31, use a multi-modal Transformer model to generate a feature vector by vector splicing fusion of audio feature vector and difficulty parameter vector ;

[0086] S32, use the Transformer encoder to process the fused feature vector x to generate the context representation h;

[0087] S33, use the Transformer decoder to generate the next note based on the context representation and the previously generated note sequence, and finally form the dynamic electronic piano score.

[0088] In S4, according to the electronic piano score, the light system on the electronic piano is used to indicate the keys that need to be played, which specifically includes: setting the light brightness proportional to the note intensity; setting the light color , where , represents the note interval time, is the color mapping function.

[0089] In S7, the user's performance audio is recorded, and a convolutional neural network is used to play the audio signal, which is compared with the standard score to evaluate the pitch and rhythm, and generate improvement suggestions, which are expressed as: pitch evaluation ; rhythm evaluation , represents the user's performance pitch, represents the standard pitch, represents the user's performance rhythm, represents the standard rhythm, and the comprehensive score , where , represents the weight, represents the tolerance threshold, based on the deviation result and the comprehensive score, a rule engine is used to generate personalized suggestions.

[0090] In one specific embodiment, the following steps are taken:

[0091] Input and Initialization: User uploads music files or records audio, and selects difficulty level, system decodes audio and initializes difficulty parameters.

[0092] Feature Extraction: Analyze music using Music Transformer, extract melody, chord and rhythm features.

[0093] Score and Guidance Generation: Fuse music features and difficulty parameters, generate dynamic score and light control signals.

[0094] Performance and Evaluation: User performs according to guidance, difficulty score and feedback module records audio, evaluates pitch and rhythm through CNN, and generates improvement suggestions.

[0095] Personalized Adjustment: User selects accompaniment or style, uses Transformer to generate personalized accompaniment content.

[0096] A specific example is provided below:

[0097] Application Scenario: User uploads music files (such as MP3, WAV format) or records audio through microphone, and selects learning difficulty according to their skill level (e.g. beginner, intermediate, advanced).

[0098] User uploads music files or records audio through user interface (UI), system uses audio processing technology to decode input. For example, use open source library (such as Librosa) to convert audio file to analyzable waveform data, ensure high-fidelity reading.

[0099] Provide difficulty selection interface, after user selection, difficulty level is mapped to quantitative parameters:

[0100] Note density (N): ): Represents the number of notes per unit time, notes per second.

[0101] Rhythm complexity (R): ): Reflects the complexity of rhythm changes.

[0102] Optimization example as follows:

[0103] Beginner: , ;

[0104] Intermediate: , ;

[0105] Advanced: , .

[0106] The teaching process is initialized according to the user's selected difficulty, ensuring that the generated music scores and guidance content match the user's needs. Through audio decoding and difficulty preset functions, the system supports multiple audio input formats and adjusts teaching content according to user skill levels, improving the personalization and adaptability of learning experience.

[0107] Intelligent analysis of user input music, extracting melody, chord, and rhythm, etc. key features, providing data support for subsequent music score generation.

[0108] Audio track separation and melody extraction: using artificial intelligence technology (Music Transformer model) to process audio data, through self-attention mechanism to analyze audio time sequence characteristics, converting input audio to MIDI format, extracting main melody note sequence.

[0109] Chord analysis: using convolutional neural network (CNN) model (Chord Recognition CNN) to identify chords in audio. For example, the user uploads an MP3 file, the system separates the main melody and accompaniment, and outputs the chord sequence (such as C-G-Am-F).

[0110] Rhythm analysis: through Onset Detection algorithm to detect note starting point, calculate rhythm pattern, generate rhythm feature vector.

[0111] Through deep learning technology, the system can efficiently and accurately extract music features, laying a foundation for dynamic music score generation and keyboard guidance.

[0112] Dynamic music score generation and keyboard highlight guidance, according to the extracted music features and user set difficulty, generate dynamic music score, and through the light system on the electronic keyboard, real-time guide user to play.

[0113] Multi-modal fusion: using multi-modal Transformer model, fusing music feature vector (including melody, chord, rhythm) and difficulty parameter , generating comprehensive feature vector .

[0114] Music score generation: Transformer encoder processes input , generates context representation : , decoder generates note sequence based on , supports multiple difficulty versions: ;

[0115] Simplified version: reduce the number of chords, reduce the complexity of rhythm.

[0116] Standard: Preserves original melody and chords.

[0117] Advanced: Adds embellishments and complex rhythms.

[0118] Key highlight guidance: Light intensity: where is the force of the th note (range [0, 127], based on MIDI standard), is the scaling factor (e.g., k=0.01, maps force to intensity range [0, 1]),

[0119] Light color: varies according to tempo , defined as the reciprocal of the interval time between adjacent notes: , the color mapping function is a piecewise function:

[0120] ,

[0121] where is in "notes per second".

[0122] Through multi-modal fusion and light control technology, generate music scores matching user difficulty, and provide intuitive key guidance, improving teaching interactivity and accuracy.

[0123] Play and evaluate, users play the electronic keyboard according to the music scores and light guidance, the system records the playing audio and evaluates the performance, providing improvement suggestions.

[0124] Collect user playing audio, use CNN model to analyze pitch and rhythm: pitch evaluation: calculate the frequency deviation of user playing notes and standard notes: Errorpitch=∣fuser−fstandard∣, rhythm evaluation: calculate the starting time deviation of notes: Errorrhythm=∣tuser−tstandard∣, comprehensive score: where and are weights, is the tolerance threshold. According to the score results, generate feedback such as "pay attention to the pitch in the high pitch area" or "strengthen rhythm stability".

[0125] Dynamic adjustment: if the user makes continuous mistakes , the system automatically switches to a simplified version of the music score.

[0126] Real-time assessment and feedback mechanism helps users identify problems and improve playing skills in a timely manner, improving learning efficiency.

[0127] Personalized arrangement, users can choose accompaniment types (such as guitar, drums) or music styles (such as pop, jazz), and generate personalized accompaniment content. Using music generation Transformer technology, conditional generation is supported.

[0128] Accompaniment generation: input: main melody and style tags ; output: accompaniment note sequence Model architecture: Transformer encoder processes main melody, decoder generates accompaniment.

[0129] Style transfer: control generation style through style embedding vector .

[0130] Conditional generation: , personalized arrangement function enhances the creative flexibility of the system and improves the interest of users in learning and playing.

[0131] The embodiments of the present application are based on artificial intelligence and multi-modal fusion technology, and the implementation principle, logic and transmission mode are as follows:

[0132] Use deep learning models (such as Transformer and CNN) to process audio data, extract music features and generate music scores and guide signals. The system quantifies difficulty parameters and performance through mathematical modeling, ensuring the accuracy and consistency of the output.

[0133] Adopt modular design, including music input, feature extraction, score generation, performance evaluation and personalized arrangement five core modules, each module in order to cooperate, form a closed loop teaching process.

[0134] Users input music and preferences through UI, the system processes and interacts with users through dynamic music scores and piano key lights, and further optimizes the guide after users play through audio feedback, realizing the teaching experience of man-machine cooperation.

[0135] The preset parameters (such as weight coefficients, difficulty thresholds, etc.) in the above formula are set and optimized by a person skilled in the art according to actual application requirements and music characteristics to ensure the accuracy and practicality of the system function. The above content demonstrates the basic principles, main features, and advantages of the present application. Those skilled in the art should understand that the present application is not limited by the above examples, and the examples and descriptions in the specification are only to illustrate the principles of the technical solutions of the present application. Without departing from the spirit and scope of the present application, the present application can also be changed and improved in various ways, such as adjusting the music analysis algorithm, expanding the light guide mode, or optimizing the personalized arrangement function, etc. These changes and improvements all fall within the scope of the present application. The scope of the present application is defined by the appended claims and their equivalents.

[0136] Another embodiment is used to illustrate an electronic piano dynamic score generation and key highlighting guide system, as shown in Figure 4 As shown, the system includes: a music input module 410 for receiving user input music and difficulty parameters and pre-processing; a feature extraction module 420 for extracting music features including melody, chord, and rhythm from pre-processed music using deep learning-based multi-track separation and chord inference algorithms; a dynamic score generation module 430 for generating dynamic electronic piano scores based on the music features and difficulty parameters using a multi-modal Transformer model, and adjusting the score content according to the note density and rhythm complexity; a key highlighting guide module 440 for indicating the keys to be played on the electronic piano using the light system on the electronic piano according to the electronic piano score; a difficulty adaptation module 450 for evaluating user skill level by collecting user key speed and accuracy, and adjusting the electronic piano score difficulty in real time according to the evaluation results; a personalized arrangement module 460 for generating accompaniment note sequences using music generation Transformer technology, and providing various accompaniment instrument options and style adjustment options according to user preferences.

[0137] In addition to the above modules, the system can also include other components, however, since these components are not related to the content of the present disclosure embodiment, their illustrations and descriptions are omitted here.

[0138] The other specific working process of the electronic piano dynamic score generation and key highlighting guide system is described with reference to the description of the above electronic piano dynamic score generation and key highlighting guide method embodiment, and will not be repeated here.

[0139] Another embodiment is used to illustrate that the system of the present application can also be implemented by means of Figure 5 the architecture of a computing device as shown. Figure 5 The architecture of this computing device is shown. As Figure 5As shown, the computer system 510 includes a system bus 530, one or more CPUs 540, input / output 520, and memory 550. The memory 550 can store various data or files used for computer processing and / or communication, as well as program instructions executed by the CPU, including the method for generating dynamic musical scores and highlighting keys for an electronic keyboard as described in this embodiment. Figure 5 The architecture shown is merely exemplary and should be adjusted according to actual needs when implementing different devices. Figure 5 One or more components in the system. The memory 550, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the electronic keyboard dynamic music score generation and key highlighting guidance method in this embodiment of the invention (e.g., the music input module 410, feature extraction module 420, dynamic music score generation module 430, key highlighting guidance module 440, difficulty adaptation module 450, and personalized arrangement module 460 in the electronic keyboard dynamic music score generation and key highlighting guidance system). One or more CPUs 540 execute various functional applications and data processing of the system of the present invention by running the software programs, instructions, and modules stored in the memory 550, that is, implementing the above-mentioned electronic keyboard dynamic music score generation and key highlighting guidance method, the method including the following steps:

[0140] Receive music and difficulty parameters input by the user and perform preprocessing;

[0141] A deep learning-based multi-track separation and chord inference algorithm is used to extract musical features, including melody, chords and rhythm, from the pre-processed music.

[0142] A multimodal Transformer model is used to generate dynamic electronic keyboard scores based on the musical features and difficulty parameters, and the score content is adjusted according to note density and rhythmic complexity.

[0143] According to the electronic keyboard sheet music, use the lighting system on the electronic keyboard to indicate the keys that need to be played;

[0144] The user's skill level is assessed by collecting the user's key press speed and accuracy, and the difficulty of the electronic keyboard sheet music is adjusted in real time based on the assessment results;

[0145] It uses music generation Transformer technology to generate accompaniment note sequences and provides a variety of accompaniment instrument options and style adjustment options according to user preferences.

[0146] Of course, the processor of the server provided in the embodiments of the present invention is not limited to performing the method operations described above, but can perform related operations in the electronic keyboard dynamic music score generation and key highlighting guidance method provided in any embodiment of the present invention.

[0147] The memory 550 can include a program storage area and a data storage area, where the program storage area can store an operating system, application programs, etc. required by at least one function; the data storage area can store data created by the use of the terminal, etc. In addition, the memory 550 can include a high-speed random access memory, and can further include a nonvolatile memory such as at least one disk memory device, a flash memory device, or other nonvolatile solid state memory device. In some examples, the memory 550 can further include a memory disposed remotely from the one or more CPUs 540, which can be connected to the device through a network. Examples of the above network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0148] The input / output 520 can be used to receive input digital or character information, and to generate key signal inputs related to user settings and function controls of the device. The input / output 520 can also include a display device such as a display screen.

[0149] The electronic piano dynamic music score generation and key highlighting guidance method provided by the embodiments of the present application can be implemented by using a computer program, and the computer program can be stored in a computer readable storage medium. The computer readable storage medium can be a volatile memory or a non-volatile memory, or a combination thereof. The non-volatile memory can be a ROM, a magnetic disk storage device, a flash memory device, or a combination thereof. The computer readable storage medium can be a tangible medium, and the tangible medium can be a computer disk, a magnetic disk, a USB flash disk, or a combination thereof.

[0150] The computer readable signal medium can include a data signal propagated in a baseband or propagated as a carrier wave in a propagated data signal, where the propagated data signal carries the computer readable program code. The propagated data signal can take various forms, including but not limited to electro-magnetic, optical, or any suitable combination thereof. The computer readable signal medium can also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate or transport program for use by or in connection with an instruction execution system, apparatus, or device.

[0151] The program code contained on a storage medium can be transmitted in any suitable form, including but not limited to electrical, optical, acoustical, or any combination of these. The program code can be transmitted from a single location, from multiple locations in a single transmission, or downloaded from different locations in multiple transmissions.

[0152] In addition, other specific working processes of the non-temporary computer readable storage medium are described with reference to the above description of the electronic piano dynamic music score generation and piano key highlighting guidance method embodiment, and will not be described again.

[0153] In this document, the terms "comprise", "contain", or any other variant thereof are intended to cover a non-exclusive inclusion, so that a process, method, comprising a series of elements not only includes those elements, but also includes other elements not explicitly listed, or further includes elements inherent to such a process, method.

[0154] The above is a further detailed description of the present application in combination with specific preferred embodiments, and cannot be considered as limiting the specific implementation of the present application to these descriptions. For ordinary skilled persons in the technical field to which the present application belongs, a number of simple deductions or replacements can be made without departing from the concept of the present application, and all of them should be considered as falling within the protection scope of the present application.

Claims

1. A method for generating dynamic musical scores and guiding key highlighting on an electronic keyboard, characterized in that, The method includes the following steps: Receive music and difficulty parameters input by the user and perform preprocessing; A deep learning-based multi-track separation and chord inference algorithm is used to extract musical features, including melody, chords and rhythm, from the pre-processed music. A multimodal Transformer model is used to generate dynamic electronic keyboard scores based on the musical features and difficulty parameters, and the score content is adjusted according to note density and rhythmic complexity. Based on the electronic keyboard sheet music, the lighting system on the keyboard is used to indicate the keys that need to be played. Specifically, this includes: setting the light brightness to be proportional to the note's dynamics; and setting the light color. in , Indicates the time interval between notes. This is a color mapping function; The user's skill level is assessed by collecting the user's key press speed and accuracy, and the difficulty of the electronic keyboard sheet music is adjusted in real time based on the assessment results; It uses Transformer technology to generate accompaniment note sequences and provides multiple accompaniment instrument options and style adjustment options based on user preferences; The difficulty parameters include note density and rhythmic complexity. Based on the thresholds of note density and rhythmic complexity, segments that meet the conditions are extracted from the analyzed music. Using a multimodal Transformer model, dynamic electronic keyboard scores are generated based on the musical features and difficulty parameters, including the following steps: A multimodal Transformer model is used to generate a feature vector x by concatenating and fusing the audio feature vector m and the difficulty parameter vector d. The fused feature vector x is processed using a Transformer encoder to generate a context representation h. The Transformer decoder is used to generate the next note based on the context representation and the previously generated note sequence, ultimately forming a dynamic electronic keyboard score. By recording user performance audio and using a convolutional neural network to compare the audio signal with standard sheet music, the system evaluates pitch and rhythm, and generates improvement suggestions. The specific expression is: Pitch Evaluation. Rhythm assessment , This indicates the notes played by the user. Indicating standard musical scales, Indicates the rhythm of the user's performance. Indicates standard rhythm, overall score ,in, , Indicates weight, The tolerance threshold is represented, and personalized recommendations are generated using a rule engine based on the deviation results and the overall score.

2. The method for generating dynamic musical scores and guiding key highlighting on an electronic keyboard according to claim 1, characterized in that, The method also includes recording user performance audio, comparing it with standard sheet music, evaluating pitch and rhythm, and generating improvement suggestions.

3. The method for generating dynamic musical scores and guiding key highlighting on an electronic keyboard according to claim 1, characterized in that, The extraction of musical features, including melody, chords, and rhythm, from the preprocessed music includes the following steps: The Music Transformer, based on a Transformer pre-trained model, is used to process audio temporal data using a self-attention mechanism. The Melody RNN model is used to convert the audio to MIDI format and extract the main melody note sequence. The Chord Recognition CNN model was used to analyze chords in audio. The rhythm pattern is calculated by detecting the starting point of the note using the Onset Detection algorithm.

4. A dynamic music score generation and key highlighting guidance system for electronic keyboards, characterized in that, The system includes: The music input module is used to receive music and difficulty parameters input by the user and to preprocess them; The feature extraction module is used to extract musical features, including melody, chords and rhythm, from the preprocessed music using a deep learning-based multi-track separation and chord inference algorithm. The dynamic score generation module is used to generate dynamic electronic keyboard scores based on the musical features and difficulty parameters using a multimodal Transformer model, and to adjust the score content according to note density and rhythmic complexity. The key highlighting guide module is used to indicate the keys to be played based on the electronic keyboard's sheet music, utilizing the keyboard's lighting system. Specifically, it includes: setting the light brightness to be proportional to the note's dynamics; and setting the light color. in , Indicates the time interval between notes. This is a color mapping function; The difficulty adaptation module is used to assess the user's skill level by collecting the user's key press speed and accuracy, and adjust the difficulty of the electronic keyboard score in real time based on the assessment results; The personalized arrangement module uses Transformer technology to generate accompaniment note sequences and provides a variety of accompaniment instrument options and style adjustment options based on user preferences. The difficulty parameters include note density and rhythmic complexity. Based on the thresholds of note density and rhythmic complexity, segments that meet the conditions are extracted from the analyzed music. Using a multimodal Transformer model, dynamic electronic keyboard scores are generated based on the musical features and difficulty parameters, including the following steps: A multimodal Transformer model is used to generate a feature vector x by concatenating and fusing the audio feature vector m and the difficulty parameter vector d. The fused feature vector x is processed using a Transformer encoder to generate a context representation h. The Transformer decoder is used to generate the next note based on the context representation and the previously generated note sequence, ultimately forming a dynamic electronic keyboard score. By recording user performance audio and using a convolutional neural network to compare the audio signal with standard sheet music, the system evaluates pitch and rhythm, and generates improvement suggestions. The specific expression is: Pitch Evaluation. Rhythm assessment , This indicates the notes played by the user. Indicating standard musical scales, Indicates the rhythm of the user's performance. Indicates standard rhythm, overall score ,in, , Indicates weight, The tolerance threshold is represented, and personalized recommendations are generated using a rule engine based on the deviation results and the overall score.

5. An electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the electronic keyboard dynamic score generation and key highlighting guidance method as described in any one of claims 1 to 3.

6. A non-transitory computer-readable storage medium, wherein computer instructions are stored on the storage medium, characterized in that, When the instruction is executed by the processor, it implements the steps of the electronic keyboard dynamic score generation and key highlighting guidance method as described in any one of claims 1 to 3.

Images