Musical Score Creation Device Using Trained Machine Learning Model
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies for creating musical scores struggle to incorporate attribute information from MIDI data, making it difficult to generate practical musical scores.
Innovation Solution
A musical score creation device and method that utilize a trained machine-learning model to estimate note and attribute information, converting input note token sequences into musical score token sequences, and then generate image musical scores, including attribute information such as key signatures, clefs, and time signatures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional MIDI data processing methods are used to create musical scores, then the creation process is simple, but attribute information cannot be estimated making the scores impractical
Solution Approach 1:
The patent introduces a trained machine learning model as an intermediary between MIDI data and musical score generation. This model estimates attribute information (tempo, dynamics, articulation) that cannot be directly obtained from MIDI data, thereby improving the practicality of generated scores without requiring complex manual annotation processes
Solution Approach 2:
The patent applies preliminary action by pre-training a machine learning model on labeled musical data before actual score generation. The model learns to estimate attribute information in advance, so that when creating new musical scores from MIDI data, the attribute estimation happens automatically without requiring real-time complex processing or manual intervention
2Measurement precision
If manual creation of musical scores with attribute information is performed, then accurate attribute information can be obtained, but the creation time and labor increase significantly
Solution Approach 1:
The patent implements self-service by enabling the system to automatically estimate attribute information using the trained machine learning model. Instead of requiring manual annotation of each musical attribute, the system autonomously generates tempo markings, dynamics, articulation symbols, and other attribute information based on the input MIDI data and learned patterns from training data
Solution Approach 2:
The patent replaces the mechanical manual process of score creation with an automated machine learning-based system. The trained model substitutes human experts in estimating attribute information, transforming a labor-intensive manual task into an automated computational process that maintains accuracy while dramatically reducing time requirements
Data Source
AI summary
A musical score creation device includes at least one processor configured to execute a receiving unit configured to receive a note sequence that includes a plurality of musical notes, and an estimation unit configured to, by using a trained model, estimate each note and attribute information for creating a musical score. The trained model is a machine-learning model that has learned an input-output relationship between a reference note sequence including a plurality of reference notes, and each reference note and reference attribute information for creating a reference musical score.