smarty4covid
收藏资源简介:
smarty4covid数据集是由雅典国立技术大学创建的一个多模态框架,用于可解释分析音频信号。该数据集包含18,265条音频记录,包括咳嗽、呼吸和语音,以及用户自报告的与COVID-19相关的信息。数据集通过众包方式收集,并经过清理和专家标注。该数据集的应用领域主要集中在开发COVID-19风险检测模型,旨在从音频记录中提取临床相关的呼吸指标,并识别咳嗽、呼吸和语音段,以支持快速、有效的COVID-19检测。
The smarty4covid dataset is a multimodal framework developed by the National Technical University of Athens for interpretable audio signal analysis. It contains 18,265 audio recordings covering cough, breathing and speech sounds, along with user-reported COVID-19-related information. The dataset was collected via crowdsourcing, and underwent cleaning and expert annotation. Its core applications focus on developing COVID-19 risk detection models, which are designed to extract clinically relevant respiratory indicators from audio recordings, identify cough, breathing and speech segments, and support rapid and effective COVID-19 testing.




