Summary of Machine Translation Advancements Of Low-resource Indian Languages by Transfer Learning, By Bin Wei et al.
Machine Translation Advancements of Low-Resource Indian Languages by Transfer Learning
by Bin Wei, Jiawei Zhen, Zongyao Li, Zhanglin Wu, Daimeng Wei, Jiaxin Guo, Zhiqiang Rao, Shaojun Li, Yuanchang Luo, Hengchao Shang, Jinlong Yang, Yuhao Xie, Hao Yang
First submitted to arxiv on: 24 Sep 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper introduces a submission by Huawei Translation Center (HW-TSC) to the WMT24 Indian Languages Machine Translation (MT) Shared Task. The authors employ two distinct knowledge transfer strategies to develop reliable machine translation systems for low-resource Indian languages. They fine-tune existing open-source models, such as IndicTrans2, and train multilingual models using bilingual data from four language pairs, including Assamese, Manipuri, Khasi, and Mizo. The results show impressive BLEU scores of up to 47.9 for bidirectional translation between English and certain languages. The authors’ approach highlights the effectiveness of transfer learning techniques for low-resource languages and contributes to advancing machine translation capabilities. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about making it easier to translate texts from different Indian languages into English and vice versa. To do this, the researchers used two methods to train their machine translation system. They first took an existing model designed for other Indian languages and adjusted it to work with Assamese and Manipuri. Then, they trained a new model using data from four language pairs, including Khasi and Mizo. The results are impressive, showing that the translations are of high quality. |
Keywords
» Artificial intelligence » Bleu » Transfer learning » Translation
