EVALUATION SUITE FOR UZBEK SPEECH AND BILINGUAL TRANSLATION AT SCALE
Keywords:
Uzbek ASR, bilingual translation, evaluation metrics, code-switching, Latin/Cyrillic, SacreBLEU, chrF, COMET.Abstract
In this article we present a reproducible evaluation suite for Uzbek speech and bilingual translation at scale. It unifies ASR, TTS, MT, and speech-to-text evaluation across Latin/Cyrillic scripts, dialectal variation, and code-switching. We provide task-specific metrics, error taxonomies and human protocols, and we report baseline scores and reliability to support fair, longitudinal benchmarking.
Downloads
Published
2025-09-23
Issue
Section
Articles
License

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
How to Cite
EVALUATION SUITE FOR UZBEK SPEECH AND BILINGUAL TRANSLATION AT SCALE. (2025). American Journal of Research in Humanities and Social Sciences, 40, 5-7. https://americanjournal.org/index.php/ajrhss/article/view/3095