Music Accompaniment Generation Using Quantized Beat and Chord Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing deep learning-based methods for generating musical accompaniment rely excessively on precise correspondence between sample music and accompaniment, ignoring inherent music theory information, leading to inaccuracies in musical style and context.
Innovation Solution
A method that quantizes musical beats and chords, extracting beat data information, chord data information, and melody data information to generate musical accompaniment data, incorporating these elements into the accompaniment melody definition, thereby improving precision and stability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If deep learning-based methods are used for accompaniment generation, then the model can learn to generate accompaniment from sample music, but the method ignores inherent music theory information and produces inaccurate musical style
Solution Approach 1:
The patent segments the accompaniment generation process into distinct modules: beat information extraction, chord information extraction, and melody generation. Each module handles specific music theory aspects separately, allowing precise control over musical style while maintaining automation. The beat and chord information serve as structured constraints that guide the generation process.
Solution Approach 2:
The patent introduces beat information and chord information as intermediary elements that mediate between the input music and the generated accompaniment. These intermediaries encode music theory knowledge and transmit it through the generation process, ensuring musical style accuracy while enabling automatic generation.
2Reliability
If sample music and accompaniment correspondence is emphasized, then the model learns generation patterns, but the musical context accuracy deteriorates
Solution Approach 1:
The patent changes the parameters used for generation from raw audio samples to structured music theory parameters (beat information, chord information, melody information). This transformation maintains generation consistency while improving musical context accuracy, as the structured parameters explicitly encode musical relationships.
3Manufacturing precision
If music parameters are integrated into accompaniment definition, then the generation detail and precision improve, but the system complexity increases
Solution Approach 1:
The patent divides the complex task of accompaniment generation into segmented processing steps: extracting beat information, extracting chord information, and generating melody information. Each segment handles a specific aspect of music theory, making the overall system more manageable and interpretable while achieving high precision.
Solution Approach 2:
The patent creates a multi-functional generation system that simultaneously processes beat information, chord information, and melody information. This universal approach handles multiple music theory aspects within a unified framework, achieving comprehensive precision without proportionally increasing complexity.
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
A music accompaniment generation method and apparatus, a device, a storage medium and a program product, relating to the technical field of computers. The method comprises: acquiring music data (210); extracting beat data information, chord data information and melody data information from the music data (220); generating music accompaniment data on the basis of the melody data information, the beat data information and the chord data information (230); and executing audio data rendering on the basis of the music accompaniment data to obtain a music accompaniment corresponding to the music data (240). In this way, music beats and music notes can be quantized, and the beats and chords are incorporated into an accompaniment melody generation process within the limitation of the music melody, so that music accompaniment data can be presented more precisely by means of music parameters, thereby improving the generation stability and the generation effect of music accompaniments. The present application can be applied to various scenarios such as music accompaniment generation.