Tune In: The AFRL WMT21 News-Translation Systems

Abstract

This paper describes the Air Force Research Laboratory (AFRL) machine translation (MT) systems and the improvements that were developed during the WMT21 evaluation campaign. We explored various methods of adapting our baseline models from WMT20 and again measure improvements in performance on the Russian English language pair.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jul 01, 2021
Accession Number
AD1157277

Entities

People

  • Grant Erdmann
  • Jeremy Gwinnup
  • Timothy Anderson

Organizations

  • 711th Human Performance Wing

Tags

Communities of Interest

  • Human Systems

DTIC Thesaurus Topics

  • Air Force
  • Air Force Research Laboratories
  • Algorithms
  • Automated Speech Recognition
  • Computational Linguistics
  • Data Sets
  • Governments
  • Language
  • Linguistics
  • Machine Translation
  • Military Research
  • Motor Skills
  • Natural Language Processing
  • Natural Languages
  • Recognition
  • Test Sets
  • Translations

Readers

  • Computational Linguistics
  • Instructional Design and Training Evaluation.
  • Materials Science and Engineering.

Technology Areas

  • AI & ML
  • AI & ML - Machine Translation
  • AI & ML - Neural Networks