Summary of Arzen-llm: Code-switched Egyptian Arabic-english Translation and Speech Recognition Using Llms, by Ahmed Heakl et al.
ArzEn-LLM: Code-Switched Egyptian Arabic-English Translation and Speech Recognition Using LLMs
by Ahmed Heakl, Youssef Zaghloul, Mennatullah Ali, Rania Hossam, Walid Gomaa
First submitted to arxiv on: 26 Jun 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Machine Learning (cs.LG)
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 investigates the challenges of machine translation (MT) and automatic speech recognition (ASR) in code-switched Egyptian Arabic-English language scenarios. The authors develop systems that translate code-switched text to either English or Egyptian Arabic, utilizing large language models like LLama and Gemma. For ASR, they employ the Whisper model for recognizing code-switched Egyptian Arabic, detailing data preprocessing and training techniques. By integrating ASR with MT, they aim to overcome limitations posed by limited resources and unique dialect features. Evaluation against established metrics shows promising results, achieving a significant improvement of 56% in English translation and 9.3% in Arabic translation. The authors highlight the importance of handling code-switching for seamless interaction in various domains, including business, culture, and academia. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at how computers can understand and translate spoken languages when people switch between two languages, like Egyptian Arabic and English. The researchers built systems that can translate this mixed language to either English or Egyptian Arabic. They used big language models called LLama and Gemma, and another model called Whisper for recognizing Egyptian Arabic. Their goal is to make computers better at understanding mixed languages so we can communicate more easily in different situations. |
Keywords
» Artificial intelligence » Llama » Translation
