linagora/linto-dataset-audio-ar-tn

LinTO阿拉伯突尼斯语音频数据集是一个包含多种音频内容（如音乐、纪录片、播客等）和对应转录文本的多样化集合。该数据集主要专注于支持突尼斯方言的语音识别任务，并包含一些法语和英语的代码转换实例。数据集分为多个配置和分割，以适应不同的实验设置，对自动语音识别（ASR）研究和开发具有很高的价值。

The LinTO DataSet Audio for Arabic Tunisian is a diverse collection of audio content including music, documentaries, podcasts, and other types of recordings, along with their corresponding transcriptions. The dataset is primarily focused on supporting speech recognition tasks for the Tunisian dialect, with some instances of code-switching between Tunisian Arabic, French, and English. It is organized into multiple configurations and splits for different experimental setups, making it valuable for Automatic Speech Recognition (ASR) research and development.