Summary of Covoswitch: Machine Translation Of Synthetic Code-switched Text Based on Intonation Units, by Yeeun Kang
CoVoSwitch: Machine Translation of Synthetic Code-Switched Text Based on Intonation Units
by Yeeun Kang
First submitted to arxiv on: 19 Jul 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI); Audio and Speech Processing (eess.AS)
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 The paper proposes a novel approach to multilingual code-switching research by creating a large-scale dataset, CoVoSwitch, which combines speech-to-text translation data from 13 languages. The authors fine-tune OpenAI’s Whisper model using PSST (speech segmentation) and then use this model to generate intonation units that are replaced with text translations from CoVoST 2. Two multilingual translation models, M2M-100 418M and NLLB-200 600M, are evaluated on the new dataset, showing improved performance in code-switching translation into English compared to monolingual settings. The study highlights language representation biases, with low-resource languages gaining most from code-switched units when translating into English, but struggling with translations into non-English. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper creates a big database of words and phrases that switch between different languages, like “Hello, how are you?” switching from English to Spanish. They use this data to test two computer programs that can translate text from one language to another. The results show that these programs do better when they’re translating into English than other languages, and that they have trouble with words that aren’t commonly used in everyday conversations. |
Keywords
» Artificial intelligence » Translation
