Conference paper Open Access

On Using SpecAugment for End-to-End Speech Translation

Bahar, Parnia; Zeyer, Albert; Schlüter, Ralf; Ney, Hermann


Dublin Core Export

<?xml version='1.0' encoding='utf-8'?>
<oai_dc:dc xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:oai_dc="http://www.openarchives.org/OAI/2.0/oai_dc/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd">
  <dc:creator>Bahar, Parnia</dc:creator>
  <dc:creator>Zeyer, Albert</dc:creator>
  <dc:creator>Schlüter, Ralf</dc:creator>
  <dc:creator>Ney, Hermann</dc:creator>
  <dc:date>2019-11-02</dc:date>
  <dc:description>This work investigates a simple data augmentation technique, SpecAugment, for end-to-end speech translation. SpecAugment is a low-cost implementation method applied directly to the audio input features and it consists of masking blocks of frequency channels, and/or time steps. We apply SpecAugment on end-to-end speech translation tasks and achieve up to +2.2% BLEU on LibriSpeech Audiobooks En→Fr and +1.2% on IWSLT TED-talks En→De by alleviating overfitting to some extent. We also examine the effectiveness of the method in a variety of data scenarios and show that the method also leads to significant improvements in various data conditions irrespective of the amount of training data.</dc:description>
  <dc:identifier>https://zenodo.org/record/3525010</dc:identifier>
  <dc:identifier>10.5281/zenodo.3525010</dc:identifier>
  <dc:identifier>oai:zenodo.org:3525010</dc:identifier>
  <dc:relation>doi:10.5281/zenodo.3525009</dc:relation>
  <dc:relation>url:https://zenodo.org/communities/iwslt2019</dc:relation>
  <dc:rights>info:eu-repo/semantics/openAccess</dc:rights>
  <dc:rights>https://creativecommons.org/licenses/by/4.0/legalcode</dc:rights>
  <dc:title>On Using SpecAugment for End-to-End Speech Translation</dc:title>
  <dc:type>info:eu-repo/semantics/conferencePaper</dc:type>
  <dc:type>publication-conferencepaper</dc:type>
</oai_dc:dc>
100
76
views
downloads
All versions This version
Views 10098
Downloads 7676
Data volume 58.4 MB58.4 MB
Unique views 8785
Unique downloads 6666

Share

Cite as