A Multimodal Dataset of Cardiac, Electrodermal, and Environmental Signals

This dataset combines signals from two domains: ambient environmental conditions (e.g., temperature, humidity, light, sound, air quality) and physiological responses (e.g., cardiac and electrodermal activity). Data was collected individually over approximately 48 minutes per participant, totaling over 600 minutes, under varied indoor conditions. Ambient signals and electrodermal activity were sampled at 1 Hz, resulting in over 30,000 data points. Cardiac activity and sound were sampled at higher frequencies, generating significantly more data points. Machine Learning models were developed using this dataset to classify ambient environmental risks to cardiac health and estimate physiological responses based solely on environmental data. The dataset enables exploration of the links between indoor climates and health, paving the way for advancements in context-aware technology.