ThBel/Accented-Speech-Fragments

Accented Speech Fragments数据集是一个大型的经过精心策划的短语音样本集合，设计用于说话人识别和对比性说话人嵌入。每个说话人都由一个唯一的说话人ID标识，并且与多个语音样本相关联。该数据集特别适合于训练对比性学习模型，目的是根据语音特征区分不同的说话人。

The Accented Speech Fragments dataset is a large curated collection of short speech samples designed for speaker identification and contrastive speaker embedding (voice printing). Each speaker in the dataset is uniquely identified by a speaker ID and is associated with multiple speech samples. This structure makes the dataset particularly suitable for training contrastive learning models, which aim to distinguish between different speakers based on their voice characteristics.