Biophysiological Signals

Two publicly available datasets, the PASS and EmpaticaE4Stress databases, were utilised in this study. They were chosen because they both used the same Empatica E4 device, which allowed the acquisition of a variety of signals, including PPG and EDA. The dataset consists of in 1587 30-second PPG segments. Each segment has been filtered and normalized using a 0.9–5 Hz band-pass and min-max normalization scheme.

A group of 10 healthy subjects without any upper limb pathologies participated in the data collection process. A total of 8 activities are performed by each subject. The measurement setup consists of a 5-channel Noraxon Ultium wireless sEMG sensor system. Representative muscle sites of the forearm are identified and self-adhesive Ag/AgCl dual electrodes are placed. The signal (sEMG) recorded during an ADL activity is segmented into functional phases: 1) rest 2) action and 3) release. During the rest phase, the subject is instructed to rest the muscles in a natural way.

This dataset comprises radar-acquired signals from 15 subjects walking on a treadmill, aimed at exploring methodologies for non-contact vital sign detection under conditions of significant body movement. Each subject participated in four experimental sessions, where radar data were collected using two Continuous Wave (CW) radars positioned to capture signals from the front and back of the subject. The data includes both raw and demodulated signals synchronized with ground-truth data obtained from a BioPac system.

The Human Activity Recognition (HAR) dataset comprises comprehensive data collected from various human activities including walking, running, sitting, standing, and jumping. The dataset is designed to facilitate research in the field of activity recognition using machine learning and deep learning techniques. Each activity is captured through multiple sensors providing detailed temporal and spatial data points, enabling robust analysis and model training.