학술논문

Predominant musical instrument detection using ensemble learning
Document Type
Conference
Source
2024 International Conference on Emerging Systems and Intelligent Computing (ESIC) Emerging Systems and Intelligent Computing (ESIC), 2024 International Conference on. :752-757 Feb, 2024
Subject
Communication, Networking and Broadcast Technologies
Components, Circuits, Devices and Systems
Computing and Processing
General Topics for Engineers
Signal Processing and Analysis
Performance evaluation
Analytical models
Instruments
Computational modeling
Simulation
Stacking
Music
Predominant musical instrument
hard voting
Soft voting
Language
Abstract
Music is the amalgamation of various instruments. It depends on the musical arrangement and composition of each music which determines the quality of the resultant music. However, there is one main instrument which is the dominant music instrument and is the backbone of the entire musical composition. In this work we combined several features of the musical sounds generated from the instrument and fed to different traditional classifiers for performance evaluation. Moreover the simulation results were also obtained by ensemble model through hard voting, soft voting and stacking model. The performance evaluation shows that Hard voting out performed all other models.