The Journal of the Acoustical Society of Korea (한국음향학회지)

The Acoustical Society of Korea (한국음향학회)
권호 표지

New Automatic Taxonomy Generation Algorithm for the Audio Genre Classification

음악 장르 분류를 위한 새로운 자동 Taxonomy 구축 알고리즘

  • 최택성 (연세대학교 전기전자공학과) ;
  • 문선국 (연세대학교 전기전자공학과) ;
  • 박영철 (연세대학교 컴퓨터정보통신공학부) ;
  • 윤대희 (연세대학교 전기전자공학과) ;
  • 이석필 (전자부품연구원(KETI) 디지털미디어 연구센터)
  • Published : 2008.04.30
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper, we propose a new automatic taxonomy generation algorithm for the audio genre classification. The proposed algorithm automatically generates hierarchical taxonomy based on the estimated classification accuracy at all possible nodes. The estimation of classification accuracy in the proposed algorithm is conducted by applying the training data to classifier using k-fold cross validation. Subsequent classification accuracy is then to be tested at every node which consists of two clusters by applying one-versus-one support vector machine. In order to assess the performance of the proposed algorithm, we extracted various features which represent characteristics such as timbre, rhythm, pitch and so on. Then, we investigated classification performance using the proposed algorithm and previous flat classifiers. The classification accuracy reaches to 89 percent with proposed scheme, which is 5 to 25 percent higher than the previous flat classification methods. Using low-dimensional feature vectors, in particular, it is 10 to 25 percent higher than previous algorithms for classification experiments.

본 논문에서는 음악 장르 분류를 위한 새로운 자동 Taxonomy 구축 알고리즘을 제안한다. 제안된 알고리즘은 모든 가능한 노드들의 분류 확률을 예측하여 예측된 분류 성능값이 가장 좋은 조합을 Taxonomy로 구축하는 것이다. 제안된 알고리즘에서의 분류 확률 예측은 훈련 데이터를 k-fold cross validation을 이용하여 분류기에 적용함으로써 이루어진다. 제안된 알고리즘을 기반으로 한 분류 성능 측정은 2 클래스로 이루어진 각각의 노드에 2개 범주 분류에 효과적인 support vector machine을 적용함으로써 이루어진다. 제안된 알고리즘의 성능 검증을 위해 음색, 리듬, 피치 등 오디오 신호의 특징을 나타내는 다양한 파라미터를 오디오 신호로부터 추출하여 제안된 알고리즘과 기존의 다중 범주 분류기들을 이용하여 분류성능을 평가하였다. 다양한 실험결과 제안된 알고리즘은 기존의 알고리즘에 비하여 5%에서 25%정도의 분류 성능이 향상된 것을 확인할 수 있었고 특히 낮은 차원의 특징벡터를 이용한 분류 실험에서는 10% 에서 25% 향상된 좋은 성능을 보였다.

Keywords

References

  1. L. Lu and H. Zhang, "Content analysis for audio classification and segmentation," IEEE Trans. on Speech and Audio Process., 10(5), 504-516, Sep. 2002 https://doi.org/10.1109/TSA.2002.804546
  2. G. Tzanetakis and P. Cook, "Musical Genre Classification of audio signals", IEEE Trans. on Speech and Audio Process., 10(4), 293-302, July 2002 https://doi.org/10.1109/TSA.2002.800560
  3. C. Yang, Database retrieval based on spectral similarity, (Stanford Univ. Database Group, Stanford, CA, Tech, Rep. 2001-14, 2001)
  4. Tao Li and Mitsunori Ogihara, "Music genre classification with taxonomy," Proc. Int. Conf. Acoustics, Speech, Signal Processing (ICASSP), 197-200, 2005
  5. Juan Jose Burred and Alexander Lerch, "A hierarchical approach to automatic musical genre classification," Proc. of the 6th Int. Conference on Digital Audio Effects (DAFX-03), London, UK, Sept. 8-11, 2003
  6. E. Scheirer and M. Slaney, "Construction and evaluation of a robust multifeature speech/music discriminator," Proc. Int. Conf. Acoustics, Speech, Signal Processing (ICASSP), 1331-1334, 1997
  7. Beth Logan, "Mel Frequency Cepstral Coefficients for music modeling," in Proc. of the First International Symposium on Music Information Retrieval (ISMIR), 2000
  8. S.Essid, G.Richard, and B.David, "Instrument Recognition in Polyphonic Music Based on Automatic taxonomies," IEEE Trans. Audio, Speech, and Lang. Process., 14(1), 68-80, Jan. 2006 https://doi.org/10.1109/TSA.2005.860351
  9. G. Peeters, "A large set of audio fetures for sound description (similarity and classification) in the CUIDADO project," CUIDADO I.S.T. Project Report, 2004
  10. D.-N. Jiang, L. Lu, H.-J. Zhang, J.-H. Tao, and L.-H. Cai, "Music type classification by spectral contrast feature,"Proc. of IEEE Int. Conf. on Multimedia and Expo (ICME02), Lausanne Switzerland, Aug, 2002
  11. S. Essid, G. Richard and B. David, "Musical instrument recognition based on class pairwise feature selection," Proc. 5th Int. Conf. Music Information Retrieval (ISMIR), Barcelona, Spain, Oct. 2004
  12. T. Tolenen and M. Karjalainen, "A computationally efficient multipitch analysis model," IEEE Trans. Speech, Audio Process, 8(6), 708-716, Nov. 2000 https://doi.org/10.1109/89.876309
  13. F. Pachet and D. Cazaly,"A taxonomy of musical genres," Proc. Content-based Multimedia Information Access (RIAO), Paris, France, 2000
  14. P. A. Devijver and J. Kitter, Pattern Recognition: A statistical approach. (New York, Prentice-Hall, 1982)
  15. J.-J. Aucouturier and F. Pachet, "Representing music genre: A state of the Art," J. of New Music Research, 32(1), 83-93, 2003 https://doi.org/10.1076/jnmr.32.1.83.16801
  16. Huan Liu and Lei Yu, "Toward integrating feature selection algorithmsfor classification and clustering," IEEE Trans. on Knowledge and Data Eng., 17(4), April 2005
  17. http://ismir2004.ismir.net/genre_contest/index.htm
  18. V. Vapnik,"The nature of statistical learning theory,"New York; Springer-Verlag, 1995
  19. D. A. Reynolds and R. C. Rose, "Robust test-independent speaker identification using Gaussian mixture speaker models," IEEE Trans. Speech, Audio Process., 3(1), 47-60, Nov. 1996
  20. S-Y. Kung and J-N. Hwang, "Neural networks for intelligent multimedia processing," Proceedingsof the IEEE, 86(6), 1244-1272, June 1998