从僧伽罗语到迪维希语：低资源语音识别的跨语言迁移学习
From Sinhala to Dhivehi: Cross-Lingual Transfer Learning for Low-Resource Speech Recognition
AI总结：
研究马尔代夫迪维希语低资源语音识别问题，通过从僧伽罗语跨语言迁移学习，采用持续预训练等五种范式实验，最强系统先在僧伽罗语预训练再在迪维希语微调，结果显示语言相关性、适应策略及解码配置对迁移学习很关键。
摘要：
马尔代夫的国语迪维希语在自动语音识别（ASR）和其他自然语言处理任务方面资源匮乏。本研究探讨从语言相关、资源相对丰富的岛屿印度 - 雅利安语僧伽罗语进行跨语言迁移学习是否能改善迪维希语ASR。进行了五种迁移学习范式的十七个实验。最强系统是先在僧伽罗语上持续预训练，再用KenLM在迪维希语上微调，其词错误率（WER）为12.89%，字符错误率（CER）为2.70%，优于仅使用迪维希语的基线。土耳其语的对照实验证实改进源于语言相关性，适应策略和解码配置也很关键。
英文摘要：
Dhivehi, the national language of the Maldives, is currently under-resourced for automatic speech recognition (ASR) and other NLP tasks. This study investigates whether cross-lingual transfer learning from Sinhala, a linguistically related, relatively well-resourced Insular Indo-Aryan language, can improve Dhivehi ASR. We conduct seventeen experiments across five transfer learning paradigms: Dhivehi-only baselines, sequential fine-tuning, multilingual fine-tuning, continual pre-training, and a control using Turkish as an unrelated language. The strongest system, continual pre-training on Sinhala followed by fine-tuning on Dhivehi with KenLM, achieves 12.89% WER and 2.70% CER, outperforming the Dhivehi-only baseline by 13.50% WER and 3.02% CER. However, the adaptation strategy and decoding configuration are equally critical for a successful transfer learning experiment. We conduct seventeen controlled experiments spanning five transfer learning paradigms: Dhivehi-only baselines, sequential fine-tuning, multilingual fine-tuning, continual pre-training, and a control experiment using Turkish as an unrelated language. The strongest system, continual pre-training on Sinhala followed by fine-tuning on Dhivehi with KenLM, achieves 12.89% WER and 2.70% CER, outperforming the Dhivehi-only baseline by 13.50% WER and 3.02% CER. The Turkish control experiment confirms that observed improvements stem from linguistic relatedness; adaptation strategy and decoding configuration are also critical.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
