2021年老人语音识别数据集

本数据库老年人说话的录音，经整理、标注等相关处理后形成，其目标是应用于老年人语音识别系统，即在不需要借助外力的作用下，依靠麦克风对老人说话信号的捕捉，实现对老年人语音信号的准确识别，进而对老人说话内容的准确理解，以更好地实现老人情感陪护。老年人语音数据库总共285097条语音数据和285097条标注文本，老年人的年龄一版在60岁以上。

This database is developed from recorded speeches of elderly individuals, following processing procedures including organization and annotation. It is targeted for applications in elderly speech recognition systems: relying solely on microphones to capture elderly speakers' vocal signals without external assistance, to achieve accurate recognition of these speech signals and precise comprehension of the spoken content, thereby enabling better emotional care for the elderly. This elderly speech database contains a total of 285,097 speech samples and 285,097 corresponding annotated transcripts, with most of the elderly subjects being 60 years old or older.