Ubenwa CryCeleb
收藏资源简介:
Ubenwa CryCeleb数据集是由Ubenwa Health创建的,包含786个新生儿的哭泣声音,总计6.5小时。该数据集的哭泣声音来自新生儿出生后和出院前的不同时间点,旨在通过验证任务推动婴儿哭泣分析的研究。数据集创建过程中,原始音频被手动分割并去除个人识别信息,确保数据的安全性和适用性。该数据集主要应用于婴儿身份验证和健康病理检测,如神经损伤等,是一个极具挑战性和应用潜力的音频数据集。
The Ubenwa CryCeleb dataset, developed by Ubenwa Health, includes crying audio recordings from 786 newborns, with a total duration of 6.5 hours. The audio samples were collected at various time points between birth and hospital discharge, and the dataset is intended to advance research on infant cry analysis via verification tasks. During the dataset's creation, raw audio was manually segmented and anonymized to remove personally identifiable information (PII), ensuring data security and usability. This dataset is mainly applied to infant identity verification and health pathology detection, such as neurological injury detection, and it is a highly challenging audio dataset with great application potential.




