zionia/isizulu-asr-1.1
收藏Hugging Face2025-11-10 更新2025-11-15 收录
下载链接：
https://hf-mirror.com/datasets/zionia/isizulu-asr-1.1
下载链接
链接失效反馈官方服务：
资源简介：
isiZulu语音识别增强数据集是一个包含isiZulu语言的语音录音和转录的数据集，针对OpenAI的Whisper自动语音识别模型进行了优化。数据集包含672个样本，语音格式为16kHz单声道WAV，最长30秒。转录文本经过清理，包括小写转换、去除标点和词性标记等处理。
The isiZulu Speech Recognition Augmented Dataset is a collection of speech recordings and transcriptions in isiZulu, optimized for OpenAIs Whisper ASR models. The dataset includes 672 samples, with audio in WAV format at 16kHz mono, up to a maximum duration of 30 seconds. The transcriptions have been cleaned, including conversion to lowercase, removal of punctuation, and POS markers.
提供机构：
zionia
