AfriNames：大多数 ASR 模型“糟蹋”非洲姓名
AfriNames: Most ASR models "butcher" African Names
AI总结：
该研究将 ASR 对非洲姓名的识别失败建模为分布偏移，并通过多语言预训练、数据增强和非洲口音微调，使相对 WER 改善 81.5%。
摘要：
有用的对话代理必须准确捕捉命名实体，以最大限度减少下游任务中的错误，例如要求语音助手播放某位艺术家的曲目、启动前往特定地点的导航，或记录患者的实验室结果。然而，当说出诸如“Ukachukwu”（伊博语）、“Lakicia”（斯瓦希里语）或“Ingabire”（卢旺达语）等命名实体时，自动语音识别（ASR）模型的性能会显著下降，并将错误传播到下游系统。我们将此问题建模为分布偏移，并证明这种模型偏差可以通过多语言预训练、提高非洲命名实体代表性的智能数据增强策略，以及在多种非洲口音上微调多语言 ASR 模型来缓解。由此得到的微调模型在包含非洲命名实体的样本上，与基线相比显示出 81.5% 的相对 WER 改善。
英文摘要：
Useful conversational agents must accurately capture named entities to minimize error for downstream tasks, for example, asking a voice assistant to play a track from a certain artist, initiating navigation to a specific location, or documenting a laboratory result for a patient. However, where named entities such as ``Ukachukwu`` (Igbo), ``Lakicia`` (Swahili), or ``Ingabire`` (Rwandan) are spoken, automatic speech recognition (ASR) models' performance degrades significantly, propagating errors to downstream systems. We model this problem as a distribution shift and demonstrate that such model bias can be mitigated through multilingual pre-training, intelligent data augmentation strategies to increase the representation of African-named entities, and fine-tuning multilingual ASR models on multiple African accents. The resulting fine-tuned models show an 81.5\% relative WER improvement compared with the baseline on samples with African-named entities.
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。
