A multi-modal wearable dataset for cognitive attention and task-based stress analysis

Stress is a natural physiological response that often adversely affects physical and mental health. Wearable devices that offer real-time acquisition of physiological signals provide opportunities for personalised monitoring and effective management of stress. We present, CATSA, a multimodal dataset comprising physiological signals, including blood volume pulse, tri-axis acceleration, electrodermal activity, and average heart rate, collected from 50 participants for monitoring different types of artificially induced stresses from wearable devices. A structured protocol was established to induce cognitive stress in participants, during which physiological signals were recorded using the wrist-worn Empatica E4 device. This dataset can potentially facilitate the development of advanced signal processing and artificial intelligence-based techniques for automated stress monitoring and contribute to the creation of intelligent systems for stress management.