Toward Accurate Music Classification Using Local Set-based Multi-label Prototype Selection

  • Authors

    • Wangduk Seo
    • Sanghyun Seo
    • Mucheol Kim
    • Jaesung Lee
    https://doi.org/10.14419/ijet.v7i3.24.24581
  • Auto Music Tags Annotation, Multi-label Learning, Prototype Selection, Pruned Problem Transformation, Local Set-based Smoother
  • Background/Objectives: Multiple music tags enable quick searching and selection of music clips for by end-users to listen to. Our goal is to improve the accuracy of automatic music categorization.

    Methods/Statistical analysis: We propose a local set-based multi-label prototype selection to remove noisy samples in datasets without transforming multi-label datasets to single-label datasets by searching the local set of each sample. To validate the superiority of proposed method, we use ten multi-label music datasets and Hamming loss as a performance measurement, which counts the symmetric difference between predicted labels and ground truth labels.

    Findings: Considering time and cost, manual categorization of a large collection of music clips is generally impractical. As such, an automated approach for addressing this task through the training of music tags annotated from an online system is employed. In the real world, multiple labels can be annotated to a music clip by users of an online system, resulting in unintended noisy samples due to inaccurate annotations. Conventional methods attempt to transform multi-label datasets to single-label datasets that can yield additional computational cost and unintended removal of non-noisy samples. In this paper, we propose a novel prototype selection method for multi-label music categorization. Experimental results indicate that the proposed method performed the best performance on nine music datasets. From the experiment of CAL500 dataset, multi-label classification performance based on the training samples selected by our proposed method was 0.1402, which indicates that 15,020 labels on average were correctly classified for 100 test samples. Compared to the second best performance by compared method, our method was able to classify 245 more labels.

    Improvements/Applications: Experimental results using ten music datasets with different subjects revealed that the proposed method yields better performance when compared to conventional methods.

     

  • References

    1. [1] Liebman, E., Saar-Tsechansky, M., & P. Stone. Dj-mc (2015). A reinforcement-learning agent for music playlist recommendation. In Proceedings of the 2015 International Conference on Autonomous Agents and Multiagent Systems, Istanbul, Turkey, pp. 591-599.

      [2] Yan, Q., Ding, C., Yin, J., & Lv., Y. (2015). Improving music auto-tagging with trigger-based context model. In Proceedings of the 2015 IEEE International Conference on Acoustics, Speech and Signal Processing, Brisbane, Australia, pp. 434-438

      [3] Bai, J., Feng, L., Peng, J., Shi, J., Luo, K., Li, Z., Liao, L., & Wang. Y. (2016). Dimensional music emotion recognition by machine learning. International Journal of Cognitive Informatics and Natural Intelligence, 10(4), 74–89.

      [4] Fu, Z., Lu, G., Ting, K. M., & Zhang, D. (2011). A survey of audio-based music classification and annotation. IEEE Transactions on Multimedia, 13(2), 303-319.

      [5] Nanni, L., Costa, Y., Lumini, A., & Kim, M. Y. (2017). Combining visual and acoustic features for music genre classification, Expert Systems with Applications, 99(1), 987-996.

      [6] Lee, J. & Kim, D.-W. (2016). Efficient Multi-Label Feature Selection Using Entropy-Based Label Selection, Entropy 18(1), 1-18.

      [7] Papanikolaou, Y., Katakis, I., & Tsoumakas, G. (2016). Hierarchical partitioning of the output space in multi-label data. Retrieved from https://arxiv.org/abs/1612.06083(arXiv:1612.06083)

      [8] Lee J. & Kim, D.-W. (2015). Memetic feature selection algorithm for multi-label classification. Information Sciences, 293(1), 80-96.

      [9] Lee J. & Kim, D.-W. (2017). SCLS: Multi-label feature selection based on scalable criterion for large label set. Pattern Recognition, 66(1), 342-352.

      [10] Zhang, Y., Gong, D.-W., Sun, X.-Y., & Y.-N. Guo. (2017). A PSO-based multi-objective multi-label feature selection method in classification. Scientific Reports, 7(376), 1-12.

      [11] Naula, P., Airola, A., Salakoski, T., and Pahikkala, T. (2014). Multi-label learning under feature extraction budgets. Pattern Recognition Letters, 40(1), 56-65.

      [12] Zhai, E., Li, Z. Li, Z. & Chen, G. (2016). Resisting tag spam by leveraging implicit user behaviors, Proceedings of the VLDB Endowment, 10(3), 241-252.

      [13] Leyva, E., Gonzalez, A. & Perez, R. (2015). Three new instance selection methods based on local sets: A comparative study with several approaches from a bi-objective perspective, Pattern Recognition, 48(1), 1523-1537.

      [14] Trohidis, K., Tsoumakas, G., Kalliris, G., & Vlahavas, I. (2008). Multi-label classification of music into emotions, In Proceedings 9th International Society Music Information Retrieval, Philadelphia, USA, pp. 325-330.

      [15] Ness, S. R., Theocharis, A., Tzanetakis, G., & Martins, L. G. (2009). Improving automatic music tag annotation using stacked generalization of probabilistic SVM outputs. In Proceedings of the 17th ACM International Conference on Multimedia, Beijing, China, pp. 705-708.

      [16] Lee, J., Jo, J.-H., Lim, H., Chae, J.-H., Lee, S.-U., & Kim, D.-W. (2015). Investigating relation of music data: Emotion and audio signals. Lecture Notes in Electrical Engineering, 330(1), 251-256.

      [17] Read, J. (2008). A pruned problem transformation method for multi-label classification. In Proceedings of New Zealand Computer Science Research Student Conference, Christchurch, New Zealand, pp. 143-150.

      [18] Lee, J. & Kim, D.-W. (2013). Feature Selection for Multi-label Classification using Multivariate Mutual Information. Pattern Recognition Letters, 34(3), 349-357.

      [19] Lartillot, O. & Toiviainen, P. (2007). A MATALB toolbox for musical feature extraction from audio. In Proceedings of the 10th International Conference on Digital Audio Effects, Bordeaux, France, pp. 237-244.

  • Downloads

  • How to Cite

    Seo, W., Seo, S., Kim, M., & Lee, J. (2018). Toward Accurate Music Classification Using Local Set-based Multi-label Prototype Selection. International Journal of Engineering & Technology, 7(3.24), 782-786. https://doi.org/10.14419/ijet.v7i3.24.24581