CLSep 22, 2021

The NiuTrans Machine Translation Systems for WMT21

arXiv:2109.10485v1649 citations
Originality Synthesis-oriented
AI Analysis

This work addresses machine translation for specific language pairs in a competition setting, but it is incremental as it builds on existing methods without introducing major innovations.

The paper describes NiuTrans's neural machine translation systems for WMT 2021, tackling multiple language directions including English-Chinese and English-Hausa, and reports results using variants of Transformer and techniques like back-translation to enhance performance.

This paper describes NiuTrans neural machine translation systems of the WMT 2021 news translation tasks. We made submissions to 9 language directions, including English$\leftrightarrow$$\{$Chinese, Japanese, Russian, Icelandic$\}$ and English$\rightarrow$Hausa tasks. Our primary systems are built on several effective variants of Transformer, e.g., Transformer-DLCL, ODE-Transformer. We also utilize back-translation, knowledge distillation, post-ensemble, and iterative fine-tuning techniques to enhance the model performance further.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes