nilc-nlp/nurc_tts_24khz
收藏资源简介:
该数据集是一个多模态语音文本数据集,包含来自巴西圣保罗和累西腓地区的音频和文本数据。数据集特征包括询问内容、片段标识、说话者信息、音频时长、转录文本、询问类型、录音年份、说话者性别、音频质量、主题和年龄组。音频采样率为24000 Hz,总大小约为83 GB,分为两个分片:圣保罗分片有120,621个样本,累西腓分片有200,295个样本,适用于语音识别、情感分析或社会语言学研究。
This dataset is a multimodal speech-text dataset containing audio and text data from São Paulo and Recife regions in Brazil. Features include inquiry content, segment identifier, speaker information, audio duration, transcribed text, inquiry type, recording year, speaker gender, audio quality, themes, and age group. The audio has a sampling rate of 24000 Hz, with a total size of approximately 83 GB, split into two parts: the São Paulo split has 120,621 samples, and the Recife split has 200,295 samples, suitable for speech recognition, sentiment analysis, or sociolinguistic research.



