mispeech/MECAT-Caption
收藏Hugging Face2025-08-13 更新2025-08-09 收录
下载链接：
https://hf-mirror.com/datasets/mispeech/MECAT-Caption
下载链接
链接失效反馈官方服务：
资源简介：
MECAT（多专家链音频任务）是一个基于大规模数据构建的全面基准，用于评估机器对音频内容的理解，包括两个核心任务：音频字幕和音频问答。
MECAT (Multi-Expert Chain for Audio Tasks) is a comprehensive benchmark constructed on large-scale data to evaluate machine understanding of audio content through two core tasks: Audio Captioning and Audio Question Answering.
提供机构：
mispeech
