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

Multi-Device and Multi-Sensor Estimation of Resting and Activity Energy Expenditure

Human energy expenditure (EE) refers to the amount of energy an individual uses to maintain essential body functions (respiration, circulation, digestion) and as a result of physical activity. EE measurement can provide valuable insights into people’s health status and help promoting a more healthy and active lifestyle. The cost and practical limitations of gold-standard EE measurement methods, led to the exploration of unobtrusive wearable sensors to estimate EE in everyday activities. However, the majority of the existing data sets used to develop such methods are not publicly available or are highly limited in terms of the type and location of sensors, which makes it difficult for the community to build on existing work and validate new methods for EE modeling.

We present WEEE, a data set collected from 17 participants to enable multimodal modeling of EE with wearable sensors.

Dataset Description

WEEE is a rich data set containing five categories of data: 1) sensor data collected using 8 wearable devices placed on 4 body positions – head, ear, chest, and wrist –, 2) respiratory gases data collected with an indirect calorimeter, which serves as ground-truth information, 3) demographics and body composition data measured using Qardio, 4) activity type – and their corresponding metabolic equivalent of task (MET) from compendium of physical activities – and intensity level, and 5) answers to questionnaires related to physical activity level, diet, stress and sleep.


Sensor Data


Device Position Sensors
VO2 Master Analyzer Face Mask Mouth Oxygen consumption
Fraction of oxygen in expired breath
Air moved by the lungs
Volume breathed in a breath
Breaths per minute
Humidity
Temperature
Pressure
Nokia Bell Labs Earbuds Ear Accelerometer
Gyroscope
Photoplethysmography
Muse S Headband Head Electroencephalography
Accelerometer
Gyroscope
Zephyr BioHarness 3.0 Chest band Chest Electrocardiography
Breathing sensor
Accelerometer
Posture sensor
Wahoo Tickr Chest-strap Chest Heart rate
Breathing sensor
Empatica E4 Wristband Wrist Accelerometer
Blood volume pulse
Electrodermal activity
Skin temperature
Apple Watch Series 6 Wrist Heart Rate
Fitbit Sense Smartwatch Wrist Heart Rate

Dataset Structure

The structure of the data is as follows:


[participant_id]/
┗ [device_id]/
┗ <data files>
Demographics.csv
Study_information.csv
Questionnaires

File Demographics.csv contains both demographics and body composition data.



Other Data


Data Type Data Tool
Demographics and Body Composition Age Qardio Scale
Gender
Height
Weight
Fat (%)
Musle (%)
Water (%)
Bone (%)
BMI
Activity Type Pen-and-paper
Intensity Level
Questionnaires Fitness International Fitness Scale (IFIS) and Physical Activity Readiness (PAR-Q)
Sleep Pittsburgh Sleep Quality Index (PSQI) and Stanford Sleepiness Scale (SSS)
Stress Perceived Stress Scale (PSS)
Diet How healthy is your diet?

Citation

If you use our dataset in your work, please cite it as follows:

Shkurta Gashi, Chulhong Min, Alessandro Montanari, Silvia Santini, & Fahim Kawsar. (2022). WEEE, A Multi-Device and Multi-Modal Dataset for Wearable Human Energy Expenditure Estimation [Data set]. Zenodo. https://doi.org/10.5281/zenodo.6420886