eduzhai > Applied Sciences > Engineering >

Surgical Mask Detection with Convolutional Neural Networks and Data Augmentations on Spectrograms

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 5 pages

Abstract: In many fields of research, labeled datasets are hard to acquire. This iswhere data augmentation promises to overcome the lack of training data in thecontext of neural network engineering and classification tasks. The idea hereis to reduce model over-fitting to the feature distribution of a smallunder-descriptive training dataset. We try to evaluate such data augmentationtechniques to gather insights in the performance boost they provide for severalconvolutional neural networks on mel-spectrogram representations of audio data.We show the impact of data augmentation on the binary classification task ofsurgical mask detection in samples of human voice (ComParE Challenge 2020).Also we consider four varying architectures to account for augmentationrobustness. Results show that most of the baselines given by ComParE areoutperformed.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×