인터넷정보학회논문지 (Journal of Internet Computing and Services)

  • 제25권4호
  • /
  • Pages.57-64
  • /
  • 2024
  • /
  • 1598-0170(pISSN)
  • /
  • 2287-1136(eISSN)
한국인터넷정보학회 (Korean Society for Internet Information)
권호 표지
DOI QR코드

DOI QR Code

음악 장르 분류를 위한 데이터 생성 및 머신러닝 적용 방안

Generating Data and Applying Machine Learning Methods for Music Genre Classification

  • 엄빛찬 ;
  • 조동휘 ;
  • 남춘성
  • Bit-Chan Eom (Department of Software Convergence Engineering, Inha University) ;
  • Dong-Hwi Cho (Department of Software Convergence Engineering, Inha University) ;
  • Choon-Sung Nam (Department of Software Convergence Engineering, Inha University)
  • 투고 : 2024.02.21
  • 심사 : 2024.07.24
  • 발행 : 2024.08.31
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

본 논문은 머신러닝을 활용하여 많은 양의 음악 데이터를 분류하여 장르 정보가 입력되어 있지 않은 음악 장르 분류 정확도 향상을 목표로 한다. 음악의 장르를 구분하기 위해 기존 연구에서 많이 사용되던 GTZAN 데이터 세트 대신 직접 데이터를 수집하고 전처리하는 방안을 제시한다. 이를 위해 GTZAN 데이터 세트보다 분류 성능이 뛰어난 데이터 세트를 생성하기 위해 Onset의 에너지 레벨이 가장 높은 일정 구간을 추출한다. 학습에 사용하는 음악 데이터의 주요 특성으로는 Mel Frequency Cepstral Coefficient(MFCC)를 포함한 57개의 특성을 이용한다. 전처리된 데이터를 통해 Support Vector Machine(SVM) 모델을 이용하여 Blues, Classical, Jazz, Country, Disco, Pop, Rock, Metal, Hiphop으로 분류한 학습 정확도가 85%를 기록하였고, 테스트 정확도가 71%를 보여주었다.

This paper aims to enhance the accuracy of music genre classification for music tracks where genre information is not provided, by utilizing machine learning to classify a large amount of music data. The paper proposes collecting and preprocessing data instead of using the commonly employed GTZAN dataset in previous research for genre classification in music. To create a dataset with superior classification performance compared to the GTZAN dataset, we extract specific segments with the highest energy level of the onset. We utilize 57 features as the main characteristics of the music data used for training, including Mel Frequency Cepstral Coefficients (MFCC). We achieved a training accuracy of 85% and a testing accuracy of 71% using the Support Vector Machine (SVM) model to classify into Classical, Jazz, Country, Disco, Soul, Rock, Metal, and Hiphop genres based on preprocessed data.

키워드

과제정보

이 논문은 정부(과학기술정보통신부)의 재원으로 정보통신기획평가원의 지원을 받아 수행된 지역지능화혁신인재양성사업(IITP-2024-RS-2023-00259678, 50%)과 정부(교육부와 한국연구재단의)재원으로 2019년 대한민국 교육부와 한국연구재단의 지원을 받아 수행된 연구임(NRF-2019S1A5A2A03040702, 50%)

참고문헌

  1. Jose M. Ausin, Enrique Bigne, Javier Marin, Jaime Guixeres and Mariano Alcaniz, "The background music-content congruence of TV advertisements: A neurophysiological study," European Research on Management and Business Economics, Vol.27, No.2, 2021. https://doi.org/10.1016/j.iedeen.2021.100154
  2. I.D Choi, "The Effect of Background Music in TV-Commercials on Brand Image: Focused on Genre of Music," The Korean Journal of Advertising and Public Relations, Vol.7, No.2, pp.35-72. 2005. https://www.dbpia.co.kr/journal/articleDetail?nodeId=NODE02499488
  3. J.I Lee and B.M Kim, "Music Genre Classification based on Musical Features of Representative Segments," Journal of KIISE:Software and Applications, Vol.35, No.11, pp.692-700, 2008. https://koreascience.kr/article/JAKO200800557077308.page
  4. Zhouyu Fu, Guojun Lu, Kai Ming Ting, and Dengsheng Zhang, "A survey of audio-based music classification and annotation," IEEE Transactions on Multimedia, Vol.13, No.2, pp.303-319, 2011. https://doi.org/10.1109/TMM.2010.2098858
  5. W.J Jang, H.W Yun, S.H Shin, H.J Cho, W. Jang, and H.C Park, "Music Genre Classification using Spikegram and Deep Neural Network," Journal of Broadcast Engineering, Vol.22, No.6, pp. 693-701, 2017. https://doi.org/10.5909/JBE.2017.22.6.693
  6. Hizlisoy, Serhat, Recep Sinan Arslan, and Emel Colakoglu, "Music Genre Recognition Based on Hybrid Feature Vector with Machine Learning Methods," Cukurova Universitesi Muhendislik Fakultesi Dergisi, Vol.38, No.3, pp.739-750, 2023. https://doi.org/10.21605/cukurovaumfd.1377737
  7. Ghildiyal, Anirudh, Komal Singh, and Sachin Sharma, "Music genre classification using machine learning," in Proc. of 2020 4th international conference on electronics, communication and aerospace technology (ICECA), IEEE, 2020. https://doi.org/10.1109/ICECA49313.2020.9297444
  8. Tarimer, Ilhan, and Buse Cennet Karadag, "Genres Classification of Popular Songs Listening by Using Keras," Gazi University Journal of Science Part A: Engineering and Innovation, Vol.11, No.1, pp.123-136, 2024. https://doi.org/10.54287/gujsa.1374878
  9. Ndou, Ndiatenda, Ritesh Ajoodha, and Ashwini Jadhav, "Music genre classification: A review of deep-learning and traditional machine-learning approaches," in Proc. of 2021 IEEE International IOT, Electronics and Mechatronics Conference (IEMTRONICS), IEEE, 2021. https://doi.org/10.1109/IEMTRONICS52119.2021.9422487
  10. Fulzele, Prasenjeet et al., "A hybrid model for music genre classification using LSTM and SVM," in Proc. of 2018 eleventh international conference on contemporary computing (IC3), IEEE, 2018. https://doi.org/10.1109/IC3.2018.8530557
  11. S.H Kim, D.S Kim, and B.W Suh, "Music Genre Classification using Multimodal Deep Learning," PROCEEDINGS OF HCI KOREA 2016, pp.389-395, 2016. https://doi.org/10.17210/hcik.2016.01.389
  12. Bob L. Sturm, "The State of the Art Ten Years After a State of the Art: Future Research in Music Information Retrieval," Journal of New Music Research, Vol.43, No.2, pp.147-172 2014. http://dx.doi.org/10.1080/09298215.2014.894533