Inferring Subjective User Evaluations From Performance Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for evaluating performance information in electronic musical instruments primarily focus on objective divergence from a correct performance, failing to infer subjective user evaluations, which are crucial for personalized user experience.
Innovation Solution
A computer-implemented method and system that utilize a trained model to infer user evaluations by processing performance information into evaluation units, leveraging machine learning to associate performance data with user feedback, enabling personalized musical lessons and preferences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If objective divergence measurement is used to evaluate performance information, then measurement precision is improved, but the ability to infer subjective user evaluation deteriorates
Solution Approach 1:
The patent introduces a trained machine learning model as an intermediary between performance information and user evaluation. The model processes objective performance data (MIDI messages, performance units) and infers subjective user evaluation attributes such as musicality, expression, and emotion. This mediator enables the system to bridge the gap between measurable objective metrics and unmeasurable subjective perceptions.
Solution Approach 2:
The patent transforms the evaluation approach by changing from fixed objective divergence metrics to dynamic subjective evaluation parameters. The machine learning model learns from training data to map performance information to evaluation attributes, allowing the system to adapt to different users' preferences and musical styles. This parameter transformation enables flexible, personalized evaluation rather than rigid standardized measurement.
2Adaptability or versatility
If machine learning model is introduced to infer user evaluation, then subjective evaluation inference capability is improved, but device complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-training the machine learning model before actual use. The model is trained on extensive training data containing performance information and corresponding user evaluations. Once trained, the model can be deployed in the electronic musical instrument to infer user evaluations without requiring complex real-time learning computations, thus reducing operational complexity while maintaining high adaptability.
Solution Approach 2:
The patent uses a trained model that replicates the complex relationship between performance information and user evaluation. Instead of implementing complex real-time analysis algorithms, the system copies the learned patterns from the trained model to quickly infer evaluations. This approach simplifies the runtime system while maintaining the capability to handle diverse user preferences and evaluation criteria.
3Productivity
If performance information is processed into evaluation units, then productivity is improved, but loss of information may occur
Solution Approach 1:
The patent segments performance information into discrete performance units (such as individual notes, chords, or musical phrases) while preserving essential characteristics. Each performance unit is processed independently through the machine learning model to generate evaluation attributes. This segmentation enables efficient parallel processing and maintains productivity while the model's trained parameters ensure that no critical information is lost during the evaluation transformation.
Data Source
AI summary
A computer-implemented method includes obtaining a trained model trained to store a relationship between first performance information and evaluation information. The first performance information includes a plurality of performance units. The evaluation information includes a plurality of pieces of evaluation information respectively associated with the plurality of performance units. The method also includes obtaining second performance information including an evaluation of each performance unit of the plurality of performance units. The method also includes processing the second performance information using the trained model to infer the evaluation of the each performance unit.