LADH (Long-term Altitude Daily Health)
收藏资源简介:
LADH数据集是由清华大学的研究团队创建的,包含来自21名参与者在五种常见个人护理场景下的240个同步RGB和红外(IR)面部视频。这些视频与真实的生理信号(如脉搏血氧饱和度、呼吸和血氧信号)同步记录,旨在解决日常护理场景中生理信号监测的挑战。该数据集的创建过程严格遵循了数据收集协议,并确保了视频与生理信号的精确同步。LADH数据集主要用于评估远程光电容积脉搏波描记术(rPPG)在真实世界环境中的鲁棒性,并探索多模态融合和多任务学习对生理信号预测的改进效果。
The LADH dataset was developed by a research team from Tsinghua University. It contains 240 synchronized RGB and infrared (IR) facial videos from 21 participants across five common personal care scenarios. These videos are synchronously recorded alongside real physiological signals including pulse oximetry, respiratory and blood oxygen signals, aiming to address the challenges of physiological signal monitoring in daily care scenarios. The construction of this dataset strictly adheres to the standardized data collection protocol, and ensures precise synchronization between the videos and the corresponding physiological signals. The LADH dataset is primarily used to evaluate the robustness of remote photoplethysmography (rPPG) in real-world environments, and to explore the improvements brought by multimodal fusion and multi-task learning for physiological signal prediction.
LADH数据集概述
数据集简介
- 全称:Non-Contact Health Monitoring During Daily Personal Care Routines
- 采集机构:青海大学
- 数据内容:240段同步非接触式面部视频(包含RGB和红外两种模态)
- 应用目标:验证rPPG技术在日常生活护理场景中的准确性和优越性
数据采集详情
- 参与者:21名(分两组:10名单日采集,11名连续10天采集)
- 采集场景:5种日常场景
- 静坐
- 坐姿刷牙梳头
- 站立
- 站姿刷牙梳头
- 运动后状态
- 生理信号:
- PPG(光电容积图)
- 呼吸频率(RR)
- 血氧饱和度(SpO2)
技术参数
- 视频参数:
- 分辨率:640×480像素
- 帧率:30 FPS
- 信号采样率:
- PPG信号:20 Hz
- 呼吸波形:50 Hz
- 采集设备:
- 面部视频:相机模块(RGB+红外)
- 生理信号:
- CMS50E脉搏血氧仪(PPG/SpO2)
- HKH-11C呼吸传感器
实验流程
- 静坐状态(2分钟):
- 佩戴呼吸传感器和血氧仪
- 保持静止注视摄像头
- 坐姿活动(2分钟):
- 保持相同设备配置
- 执行刷牙/梳头动作
- 站立状态(2分钟):
- 站立姿势采集
- 站姿活动(2分钟):
- 站立姿势执行刷牙/梳头动作
- 运动后状态(2分钟):
- 进行体能运动后采集




