Emotional Audio-Textual Depression Corpus (EATD-Corpus)
收藏资源简介:
EATD-Corpus是由同济大学创建的第一个包含中文音频和文本数据的开源抑郁症数据集,包含162名志愿者的音频和文本转录。数据集通过一个应用程序收集,志愿者需回答三个随机问题并完成SDS问卷。数据集的创建过程包括数据收集和预处理,确保音频和文本的质量。EATD-Corpus主要用于自动抑郁症检测研究,旨在通过分析语音特征和语言内容来提高抑郁症诊断的准确性。
EATD-Corpus is the first open-source depression dataset encompassing both Chinese audio and text data, developed by Tongji University. It comprises audio recordings and text transcriptions from 162 volunteers. The dataset was collected through an application, wherein volunteers were asked to answer three random questions and complete the SDS questionnaire. The creation of EATD-Corpus involves data collection and preprocessing steps to guarantee the quality of the audio and text materials. Primarily utilized for automatic depression detection research, EATD-Corpus aims to enhance the accuracy of depression diagnosis by analyzing speech features and linguistic content.




