Negative Emotion Management Using a Smart Shirt and a Robot Assistant

Abstract

Negative affects such as anger, fear, nervousness, depression, etc., may increase human's susceptibility to illness. In this letter, we propose a negative emotion management system that is able to recognize negative emotions through ECG signal and perform emotion regulation through a robot assistant, which has a potential for reducing health risks. A smart shirt is developed to collect the ECG signal from the human body. The robot assistant has the ability to engage in verbal conversations with humans. Recurrence Quantitative Analysis (RQA) is used to extract ECG features for emotion classification purpose. Along with our own dataset, two other public datasets, RECOLA and DECAF, are also used to evaluate our methodology. The detection of negative emotion can trigger the robot assistant to help the user get out of such situations through interactive conversations. We tested and evaluated the proposed framework through experiments. We also assessed the effectiveness of the interactions with the robot on the emotional well-being of older adults.