Stress and Emotion Detection in IoT Systems with Machine Learning Techniques and Wearable Devices

#IoT #AI #MachineLearning #EmotionDetection #SignalProcessing #Research #AcademicPublishing

Nowadays, due to the fast pace of society and the pressure from all aspects of work and life, many individuals suffer from stress and associated negative emotions. Prolonged exposure to such emotions can escalate into more severe mental and physical disorders, underlining the importance of monitoring emotional changes. The advent of Internet of Things (IoT) systems has opened new avenues for analyzing and predicting people's stress or emotions through technology, which can enable people to pay more attention to their mental health in time. As emotional fluctuations can affect physiological signals, monitoring these can offer insight into one's emotional state. Specifically, the fluctuations in skin conductivity resulting from different sweat secretion levels make electrodermal activity (EDA) signals a promising candidate for emotion detection. With the advantages of wearable devices equipped with EDA and other physiological signal sensors in IoT systems, there is a convenient and cost-effective avenue to collect the EDA data. This seminar will introduce the mechanisms of stress and emotion detection with EDA and methods to process and analyze the signals. Moreover, the existing difficulties and future directions in this field will also be discussed.

Furthermore, the speaker will share insights from her highly productive Ph.D. journey, particularly focusing on her extensive experience in publishing journal and conference papers.