Gamelan Demung Music Transcription Based on STFT Using Deep Learning

Hermawan, Andi Rokhman; Yuniarno, Eko Mulyanto; Wulandari, Diah Puspito

doi:10.12962/jaree.v6i2.276

Cited by 1 publication

(1 citation statement)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The separation of two or more signals is very important in finding the characteristics of the signals. In addition, the separation of identical instruments can also be used for sound recognition from musical instruments [3]. Although it is currently difficult to define  ISSN: 2252-8938 Int J Artif Intell, Vol.…”

Section: Introductionmentioning

confidence: 99%

Deep learning for audio signal-based tempo classification scenarios

Muljono,

Nurtantio Andono,

Ayu Wulandari

et al. 2024

IJ-AI

View full text Add to dashboard Cite

This article explains how to determine the tempo of the kendhang, an Indonesian traditional melodic instrument. This research presents novelty as technological research related to gamelan instruments, which has rarely been achieved thus far, through the introduction of kendhang tempo types through the sounds produced, with the hope of creating an automatic system that can recognize the kendhang tempo during a gamelan performance. The testing in this work will categorize the tempo of kendhang into three categories: slow, medium, and fast, utilizing one of the two scenario models proposed, mel frequency cepstral coefficients (MFCC) and convolutional neural network (CNN) in the first scenario, and mel spectrogram and CNN in the second. Kendhang's original audio data, which was captured in real time and later enhanced, makes up the data set. The model 1 scenario, which entails feature extraction using MFCC and classification using the CNN classification approach, is the best scenario in this research, based on the experimental results. When compared to the other suggested modeling scenarios, model 1 has a level of 97%, an average accuracy, and a gain value of 96.67%, making it a solid assistant in terms of kendhang's good tempo recognition accuracy.

show abstract

Section: Introductionmentioning

confidence: 99%