IMU-Based Reliable Vital Signs Monitoring From Human to Dog.

This paper introduces an innovative processing pipeline designed to estimate respiratory rate (RR) and heart rate (HR) using acceleration data from inertial measurement units (IMUs), addressing a critical monitoring task applicable to both humans and animals. The key features of the pipeline include an automated axis selection method and a movement-aware scheme, which collectively enhance the precision and confidence of vital sign estimations. Extensive validation of this approach was conducted across human and dog subjects, which produced highly promising results. Specifically, when applied to humans in a supine posture, the pipeline achieves an average error of 0.27 bpm for RR and 3.68 bpm for HR. In the context of sleeping dogs, the implementation demonstrated an average error of 0.3 bpm for RR and 1.8 bpm for HR. These outcomes underscore the robustness and adaptability of the proposed methodology, which offers significant potential for diverse applications across various species in different scenarios.