16th International Workshop on Spoken Language Translation 2019
Published November 2, 2019 | Version v1
Conference paper Open

The IWSLT 2019 KIT Speech Translation System

Creators

  • 1. Institute for Anthropomatics and Robotics, KIT - Karlsruhe Institute of Technology, Germany
  • 2. IInstitute for Anthropomatics and Robotics, KIT - Karlsruhe Institute of Technology, Germany
  • 3. Department of Data Science, University of Maastricht, Netherlands
  • 4. Institute for Anthropomatics and Robotics KIT - Karlsruhe Institute of Technology, Germany
  • 5. Language Technologies Institute, Carnegie Mellon University & Institute for Anthropomatics and Robotics KIT - Karlsruhe Institute of Technology, Germany

Description

This paper describes KIT’s submission to the IWSLT 2019 Speech Translation task on two sub-tasks corresponding to two different datasets. We investigate different end-to-end architectures for the speech recognition module, including our new transformer-based architectures. Overall, our modules in the pipe-line are based on the transformer architecture which has recently achieved great results in various fields. In our systems, using transformer is also advantageous compared to traditional hybrid systems in term of simplicity while still having competent results.

Files

IWSLT2019_paper_40.pdf

Files (233.7 kB)

Name Size Download all
md5:c328ea0a9f886839abd3b19cac97b40b
233.7 kB Preview Download