dsfsi/lwazi-asr-corpus-compressed

Lwazi ASR语料库是一个针对多种低资源南非语言的自动语音识别（ASR）的精选集合。这些数据集旨在用于语音识别研究和开发，特别是针对代表性不足的语言。每个语料库包含来自母语使用者的电话对话脚本录音及其对应的转录。这些音频记录主要来自一般领域的电话对话。所有数据都遵循Creative Commons BY 3.0授权，允许开放研究和带归属的非商业用途。

The Lwazi ASR Corpus is a curated collection of Automatic Speech Recognition (ASR) datasets for several low-resourced South African languages. These datasets are designed for use in speech recognition research and development, particularly for underrepresented languages. Each corpus consists of scripted telephonic speech recordings from native speakers, along with corresponding transcriptions. The audio recordings are mainly from general domain telephone conversations. All data is licensed under Creative Commons BY 3.0, allowing for open research and non-commercial use with attribution.