1
views
0
recommends
+1 Recommend
0 collections
    0
    shares
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      Learning Sound Event Classifiers from Web Audio with Noisy Labels

      Preprint

      Read this article at

      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          As sound event classification moves towards larger datasets, issues of label noise become inevitable. Web sites can supply large volumes of user-contributed audio and metadata, but inferring labels from this metadata introduces errors due to unreliable inputs, and limitations in the mapping. There is, however, little research into the impact of these errors. To foster the investigation of label noise in sound event classification we present FSDnoisy18k, a dataset containing 42.5 hours of audio across 20 sound classes, including a small amount of manually-labeled data and a larger quantity of real-world noisy data. We characterize the label noise empirically, and provide a CNN baseline system. Experiments suggest that training with large amounts of noisy data can outperform training with smaller amounts of carefully-labeled data. We also show that noise-robust loss functions can be effective in improving performance in presence of corrupted labels.

          Related collections

          Author and article information

          Journal
          04 January 2019
          Article
          1901.01189
          4d9bbe00-6ac4-4d62-a1c9-ce2f6394e098

          http://creativecommons.org/licenses/by/4.0/

          History
          Custom metadata
          cs.SD cs.LG eess.AS stat.ML

          Machine learning,Artificial intelligence,Electrical engineering,Graphics & Multimedia design

          Comments

          Comment on this article