We investigate low-resource Arabic-Asian machine translation across six Arabic-centric translation directions: Arabic↔English, Arabic↔Hindi, and Arabic↔Urdu. Fine-tuned MADLAD-400 ranks first in four of the six directions on the official primary leaderboard and places within the top three in the remaining two. We fine-tune three pretrained multilingual models, MADLAD-400, NLLB-200, and GemmaX2, and compare them with their zero-shot counterparts using BLEU, chrF2++, COMET-22, and TER. Across all language pairs, supervised fine-tuning consistently outperforms zero-shot inference. Among the evaluated models, fine-tuned MADLAD-400 achieves the strongest overall performance, NLLB-200 delivers competitive results, and GemmaX2-28-9B exhibits the largest relative improvement over its zero-shot baseline. These findings demonstrate the effectiveness of supervised adaptation for low-resource Arabic-Asian machine translation and highlight MADLAD-400 as the most robust model across the evaluated translation directions.

Cross-lingual voice cloning (CLVC) aims to synthesize speech in a target language while preserving the vocal identity of a source speaker who has no recorded speech in that language. Despite recent advances in multilingual text-to-speech systems, zero-shot CLVC remains challenging due to phonetic divergence across languages and the difficulty of maintaining speaker identity alongside linguistic intelligibility. In this work, we present a systematic evaluation of four state-of-the-art CLVC systems spanning autoregressive and diffusion-based architectures. Using English source speakers from the ACL-60/60 dataset, we evaluate zero-shot voice transfer across multiple target languages, including Arabic, Chinese, French, German, Russian, and Japanese. Systems are assessed using speaker similarity and content consistency metrics under a unified multilingual evaluation pipeline. We analyze how different modeling approaches autoregressive language modeling and diffusion-based flow matching handle the tradeoff between speech accuracy and speaker identity preservation across different architectural approaches. We further observe substantial performance variation across languages, with Arabic remaining particularly challenging under zero-shot transfer settings.