AfroDigits

AfroDigits是一个由Hugging Face团队开发的社区驱动型语音数据集，专注于非洲语言，目前覆盖38种非洲语言。该数据集包含2185个音频样本，主要用于语音处理任务，如电话号码和街道号码的识别。数据集的创建过程采用了社区参与的方法，旨在填补非洲语音语料库的空白。AfroDigits的应用领域包括教育和实际应用，如语音识别教程和模型评估。

AfroDigits is a community-driven speech dataset developed by the Hugging Face team, focusing on African languages and currently covering 38 African languages. This dataset contains 2,185 audio samples, primarily used for speech processing tasks such as the recognition of phone numbers and street numbers. The dataset was created using a community-participatory approach, aiming to fill the gap in African speech corpora. The application areas of AfroDigits include education and practical scenarios, such as speech recognition tutorials and model evaluation.