Pervasive and unobtrusive emotion sensing for human mental health

In this paper, we present a pervasive and unobtrusive system for sensing human emotions, which are inferred based on the recording, processing, and analysis of the Galvanic Skin Response (GSR) signal from human bodies. Being different from traditional multi modal emotion sensing systems, our proposed system recognizes human emotions with the single modularity of GSR signal, which is captured by wearable sensing devices. A comprehensive set of features is extracted from GSR signal and fed into supervised classifiers for emotion identification. Our system has been evaluated by specific experiments to investigate the characteristics of human emotions in practice. The high accuracy of emotion classification highlights the great potential of this system in improving humans' mental health in the future.