A multidevice and multimodal dataset for human energy expenditure estimation using wearable devices

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Human Physiology

Background:

• Accurate estimation of human energy expenditure (EE) is crucial for health monitoring and lifestyle interventions.
• Existing datasets may lack the multi-modal and multi-device diversity needed for robust EE estimation models.
• Wearable sensors offer a promising avenue for continuous and unobtrusive EE monitoring.

Purpose of the Study:

• To introduce the WEEE dataset, a novel multi-device and multi-modal collection for human energy expenditure research.
• To provide a comprehensive resource encompassing wearable sensor data, respiratory measurements, and participant information.
• To facilitate the development of advanced EE estimation techniques across various applications.

Main Methods:

• Collected data from 17 participants engaged in diverse physical activities.
• Utilized seven wearable devices across four body locations (head, ear, chest, wrist).
• Incorporated ground-truth respiratory data from an indirect calorimeter, alongside demographic and activity information.

Main Results:

• The WEEE dataset includes synchronized multi-sensor readings and validated energy expenditure metrics.
• Data encompasses wearable sensor streams, indirect calorimetry, participant demographics, activity types, and MET values.
• Questionnaire data on physical activity, diet, stress, and sleep are also integrated.

Conclusions:

• The WEEE dataset's richness and diversity are expected to spur innovation in human energy expenditure estimation.
• This resource will enable the creation of more accurate and adaptable EE monitoring systems.
• Applications range from personalized health coaching to advanced human activity recognition.