General audio tagging with ensembling convolutional neural networks and statistical features.

Journal: The Journal of the Acoustical Society of America
Published Date:

Abstract

Audio tagging aims to infer descriptive labels from audio clips and it is challenging due to the limited size of data and noisy labels. The solution to the tagging task is described in this paper. The main contributions include the following: an ensemble learning framework is applied to ensemble statistical features and the outputs from the deep classifiers, with the goal to utilize complementary information. Moreover, a sample re-weight strategy is employed to address the noisy label problem within the framework. The approach achieves a mean average precision of 0.958, outperforming the baseline system with a large margin.

Authors

  • Kele Xu
    Department of Engineering, Université Pierre et Marie Curie, Paris 75005, France kelele.xu@gmail.com.
  • Boqing Zhu
    National Key Laboratory of Parallel and Distributed Processing, National University of Defense Technology, Changsha, People's Republic of China.
  • Qiuqiang Kong
    Centre for Vision, Speech and Signal Processing (CVSSP), University of Surrey, Guildford GU2 7XH, United Kingdom.
  • Haibo Mi
    School of Computer, National University of Defense Technology, Changsha 410073, China. haibo_mihb@126.com.
  • Bo Ding
    National Key Laboratory of Parallel and Distributed Processing, National University of Defense Technology, Changsha, People's Republic of China.
  • Dezhi Wang
    College of Meteorology and Oceanography, National University of Defense Technology, Changsha, People's Republic of Chinakelele.xu@gmail.com, zhuboqing09@nudt.edu.cn, q.kong@surrey.ac,uk, mihaibo23@nudt.edu.cn, dingbo@nudt.edu.cn, wang_dezhi@hotmail.com, wanghuaimin22@nudt.edu.cn.
  • Huaimin Wang
    National Key Laboratory of Parallel and Distributed Processing, National University of Defense Technology, Changsha, People's Republic of China.