Mrinal007/mkbhd_tts_medium_clean
收藏资源简介:
这是一个音频-文本配对数据集,包含185个训练样本,每个样本由音频、文本和来源三部分组成。音频特征采样率为24000Hz,适用于语音处理和识别任务;文本特征为字符串类型,提供与音频对应的转录或描述;来源特征标识数据来源。数据集总大小为233196506字节,下载大小为233185097字节,数据文件路径为train-*。该数据集可能用于语音合成、语音识别或音频-文本对齐等自然语言处理应用。
This is an audio-text paired dataset containing 185 training samples, each consisting of audio, text, and source components. The audio feature has a sampling rate of 24000Hz, suitable for speech processing and recognition tasks; the text feature is of string type, providing transcriptions or descriptions corresponding to the audio; and the source feature identifies the data origin. The total dataset size is 233196506 bytes, with a download size of 233185097 bytes, and data files are located at train-*. This dataset is likely intended for applications such as speech synthesis, speech recognition, or audio-text alignment in natural language processing.




