hon9kon9ize/yue_emo_speech

该数据集包含从YouTube和RTHK爬取的1,000小时粤语情感语音，每个语音片段都标注了情感标签（如愤怒、厌恶、恐惧、快乐、中性、其他、悲伤、惊讶）以及情感标签的置信度。音频文件经过降噪处理，文本转录由SenseVoiceSmall生成，并使用MinHash进行去重。This dataset contains 1,000 hours of Cantonese speech crawled from YouTube and RTHK, each labeled with one of the following emotions: angry, disgusted, fearful, happy, neutral, other, sad, or surprised, along with the confidence of the emotion label. The audio files are denoised, transcriptions are generated by the SenseVoiceSmall model, and deduplicated using MinHash.