KoelLabs/speech-accent-context
收藏Hugging Face2025-03-12 更新2025-04-12 收录
下载链接：
https://hf-mirror.com/datasets/KoelLabs/speech-accent-context
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含2138名来自不同语言背景的英语发音者朗读特定句子的音频文件。这些句子中特别关注三个单词：零食(snack)、蛇(snake)和包(bags)，这些单词在脱离上下文的情况下容易混淆。数据集提供了完整的音频、 Facebook转录、以及有上下文和无上下文的单词转录。该数据集可用于研究发音识别和语言模型的能力。
The dataset includes audio recordings of 2138 English speakers with diverse linguistic backgrounds reading a specific sentence. It focuses on three words: snack, snake, and bags, which are notoriously difficult to distinguish without context. The dataset provides full audio, Facebook transcriptions, and transcriptions of the words with and without context. It can be used to study pronunciation recognition and the capabilities of language models.
提供机构：
KoelLabs
