Multidomain Mobile Video Physiology Dataset (MMPD)
收藏资源简介:
MMPD数据集由清华大学创建,包含33名不同肤色(Fitzpatrick 3-6)的参与者在四种不同光照条件和四种活动下的11小时视频记录。数据集旨在通过捕捉多样化的肤色、身体运动和光照条件,推动远程光电容积脉搏波(rPPG)技术的发展。创建过程中,使用了三星Galaxy S22 Ultra手机进行视频录制,并与HKG-07C+血氧仪同步记录生理信号。该数据集适用于开发和测试能够处理复杂现实场景的rPPG算法,特别是在远程健康监测和慢性心血管疾病管理中的应用。
The MMPD dataset was developed by Tsinghua University. It contains 11 hours of video recordings from 33 participants with diverse skin tones (Fitzpatrick scale 3-6), acquired under four distinct lighting conditions and four types of daily activities. The dataset is designed to promote the advancement of remote photoplethysmography (rPPG) technologies by capturing varied skin tones, bodily motions and lighting environments. During the dataset construction, video recordings were captured using a Samsung Galaxy S22 Ultra smartphone, and physiological signals were synchronously logged with an HKG-07C+ oximeter. This dataset is applicable for developing and testing rPPG algorithms capable of handling complex real-world scenarios, especially for applications in remote health monitoring and chronic cardiovascular disease management.




