kotoba-speech/mls-eng-full

这是一个包含音频及其相关特征的数据集，特征包括转录文本、原始文件路径、开始和结束时间、原始文本、音频持续时间、说话者ID、书籍ID、信噪比、C50值、语音持续时间、发音速率、音素、STOI值、SI-SDR值、PESQ值、平均音高、音高标准差、性别、音高、噪声、混响、语音单调性、SDR噪声、PESQ语音质量、文本描述等。数据集分为训练集、验证集和测试集，适用于语音信号处理和语音识别等领域。

This dataset consists of audio data and its related features, including transcript text, original path, timestamps, audio duration, speaker ID, book ID, signal-to-noise ratio, speech monotony, and other acoustic features. It is split into training, validation, and test sets, with detailed information on the size and number of examples for each part. The configuration section provides path information for the data files.