英国COVID-19语音音频数据集

英国COVID-19语音音频数据集是由英国卫生安全局创建的大规模PCR参考语音音频数据集，旨在用于训练和评估机器学习模型，以分类SARS-CoV-2感染状态或相关呼吸症状。数据集包含72,999名参与者的音频记录，包括自愿咳嗽、呼气和语音，并链接了SARS-CoV-2测试结果。数据集创建过程中，参与者通过国家测试和追踪计划以及REACT-1调查在英格兰从2021年3月至2022年3月被招募。数据集的应用领域包括生物声学研究和COVID-19筛查，旨在提供一种非侵入性、经济且可扩展的COVID-19监测方法。

The UK COVID-19 Speech Audio Dataset is a large-scale PCR-referenced speech audio dataset developed by the UK Health Security Agency, designed for training and evaluating machine learning models to classify SARS-CoV-2 infection status or related respiratory symptoms. The dataset contains audio recordings from 72,999 participants, including voluntary coughs, exhalations, and speech samples linked to SARS-CoV-2 test results. During the dataset's creation, participants were recruited across England between March 2021 and March 2022 via the National Test and Trace programme and the REACT-1 survey. Its application fields cover bioacoustics research and COVID-19 screening, with the goal of providing a non-invasive, cost-effective and scalable approach for COVID-19 surveillance.

搜集汇总

背景与挑战

背景概述

英国COVID-19语音音频数据集是一个大规模的声音音频数据集，用于训练和评估机器学习模型以分类COVID-19感染状态或相关呼吸道症状。数据集包含咳嗽、呼气等音频记录，并与人口统计、症状和PCR测试结果相关联，是目前最大的PCR参考音频记录集合。