eduzhai > Applied Sciences > Engineering >

They are wearing a mask Identification of Subjects Wearing a Surgical Mask from their Speech by means of x-vectors and Fisher Vectors

  • king
  • (0) Download
  • 20210507
  • Save

... pages left unread,continue reading

Document pages: 5 pages

Abstract: Challenges based on Computational Paralinguistics in the INTERSPEECHConference have always had a good reception among the attendees owing to itscompetitive academic and research demands. This year, the INTERSPEECH 2020Computational Paralinguistics Challenge offers three different problems; here,the Mask Sub-Challenge is of specific interest. This challenge involves theclassification of speech recorded from subjects while wearing a surgical mask.In this study, to address the above-mentioned problem we employ two differenttypes of feature extraction methods. The x-vectors embeddings, which is thecurrent state-of-the-art approach for Speaker Recognition; and the FisherVector (FV), that is a method originally intended for Image Recognition, buthere we utilize it to discriminate utterances. These approaches employ distinctframe-level representations: MFCC and PLP. Using Support Vector Machines (SVM)as the classifier, we perform a technical comparison between the performancesof the FV encodings and the x-vector embeddings for this particularclassification task. We find that the Fisher vector encodings provide betterrepresentations of the utterances than the x-vectors do for this specificdataset. Moreover, we show that a fusion of our best configurations outperformsall the baseline scores of the Mask Sub-Challenge.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...