WEEE, A Multi-Device and Multi-Modal Dataset for Wearable Human Energy Expenditure Estimation

We present WEEE, a multi-device and multi-modal dataset collected from 17 participants under different physical activities.
WEEE contains: 1) sensor data collected using 7 wearable devices placed on 4 body locations - head, ear, chest, and wrist
-, 2) respiratory data collected with an indirect calorimeter serving as ground-truth information, 3) demographics and body
composition data (e.g., muscle or fat percentage), 4) activity type - and their corresponding metabolic equivalent of task (MET) values - and intensity level, and 5) answers to questionnaires related to physical activity level, diet, stress and sleep. Thanks to the diversity of sensors and body locations of the WEEE dataset, we envision that this dataset will enable the development of novel human energy expenditure estimation techniques for a diverse set of application scenarios. Energy expenditure (EE) refers to the amount of energy an individual uses to maintain body functions and as a result of physical activity. The ability to estimate EE allows computing systems obtaining valuable insights regarding people’s physical activity and providing personalized recommendations for promoting a healthier and more active lifestyle.

我们提出WEEE数据集，这是一款多设备多模态数据集，采集自17名参与不同身体活动的受试者。WEEE数据集包含以下内容：1）由7台可穿戴设备采集的传感器数据，这些设备佩戴于头部、耳部、胸部与手腕4个身体部位；2）通过间接热量计采集的呼吸数据，该数据作为基准真值（ground-truth）信息；3）人口统计学与身体成分数据（例如肌肉百分比、脂肪百分比）；4）活动类型及其对应的代谢当量（metabolic equivalent of task，MET）值与活动强度等级；5）关于身体活动水平、饮食、压力与睡眠的问卷应答数据。鉴于WEEE数据集在传感器类型与佩戴部位上的多样性，我们预期该数据集将助力研发面向多元应用场景的新型人体能量消耗估算技术。能量消耗（Energy expenditure，EE）指个体用于维持身体机能以及进行身体活动所消耗的总能量。准确估算能量消耗的能力，可使计算系统获取关于个体身体活动的有效洞察，并为推广更健康、更积极的生活方式提供个性化建议。