Semi-Supervised Audio Classification with Partially Labeled Data

被引:4
作者
Gururani, Siddharth [1 ]
Lerch, Alexander [1 ]
机构
[1] Georgia Inst Technol, Atlanta, GA 30332 USA
来源
23RD IEEE INTERNATIONAL SYMPOSIUM ON MULTIMEDIA (ISM 2021) | 2021年
关键词
audio classification; semi-supervised learning;
D O I
10.1109/ISM52913.2021.00027
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
Audio classification has seen great progress with the increasing availability of large-scale datasets. These large datasets, however, are often only partially labeled as collecting full annotations is a tedious and expensive process. This paper presents two semi-supervised methods capable of learning with missing labels and evaluates them on two publicly available, partially labeled datasets. The first method relies on label enhancement by a two-stage teacher-student learning process, while the second method utilizes the mean teacher semi-supervised learning algorithm. Our results demonstrate the impact of improperly handling missing labels and compare the benefits of using different strategies leveraging data with few labels. Methods capable of learning with partially labeled data have the potential to improve models for audio classification by utilizing even larger amounts of data without the need for complete annotations.
引用
收藏
页码:111 / 114
页数:4
相关论文
共 18 条
  • [1] Akiyama O., 2019, DCASE WORKSHOP
  • [2] Bachman P, 2014, ADV NEUR IN, V27
  • [3] Berthelot D, 2019, ADV NEUR IN, V32
  • [4] Cartwright M., 2020, DCASE WORKSHOP
  • [5] Cramer J, 2019, INT CONF ACOUST SPEE, P3852, DOI 10.1109/ICASSP.2019.8682475
  • [6] Defferrard M., 2017, 18 INT SOC MUS INF R
  • [7] Fonseca E., 2020, ARXIV201000475 CS EE
  • [8] Addressing Missing Labels in Large-Scale Sound Event Recognition Using a Teacher-Student Framework With Loss Masking
    Fonseca, Eduardo
    Hershey, Shawn
    Plakal, Manoj
    Ellis, Daniel P. W.
    Jansen, Aren
    Moore, R. Channing
    [J]. IEEE SIGNAL PROCESSING LETTERS, 2020, 27 : 1235 - 1239
  • [9] Gemmeke JF, 2017, INT CONF ACOUST SPEE, P776, DOI 10.1109/ICASSP.2017.7952261
  • [10] Gururani S., 2019, ISMIR