AI Tool From UAE Achieves 97% Accuracy in Identifying Arabic Dialects
A new AI tool developed at the University of Sharjah can distinguish between 19 dialects across 22 Arabic-speaking countries.
Scientists at the University of Sharjah have developed an advanced AI system capable of recognizing Arabic dialects with precision, achieving The system addresses the complexities of the Arabic language, achieving 97.29% accuracy in identifying regional dialects and 94.92% accuracy for country-specific variations across 22 nations.
The AI model was trained on over 3,000 hours of audio sourced from YouTube, covering 19 dialects from the Middle East, North Africa, and the Arabian Peninsula. One of the key achievements of the project is its efficiency; the AI system required only 29% of the typical training data used in similar research.
Potential applications for this AI tool include virtual assistants, translation services and customer support, enhancing communication for millions of Arabic speakers across diverse regions. The tool is publicly available for further development, attracting industry-wide attention due to its potential for broad adoption in AI-driven language services.
In a related development, Saudi Arabia is collaborating with IBM's AI platform, watsonx, to create a generative AI program capable of handling Arabic dialects, signaling further progress in AI language technology in the region.
