IoT-based Remote Facial Expression Monitoring System with sEMG Signal

Biopotentials including Electrocardiography (ECG), Electromyography (EMG) and Electroencephalography (EEG) measure the activity of heart, muscles and brain, respectively. They can be used for noninvasive diagnostic applications, assistance in rehabilitation medicine and human-computer interaction. The concept of Internet of Things (IoT) can bring added value to applications with biopotential signals in healthcare and human-computer interaction by integrating multiple technologies such as sensors, wireless communication and data science. In this work, we present a wireless biopotentials remote monitoring and processing system. A prototype with the case study of facial expression recognition using four channel facial sEMG signals is implemented. A multivariate Gaussian classifier is trained offline from one person's surface EMG (sEMG) signals with four facial expressions: neutral, smile, frown and wrinkle nose. The presented IoT application system is implemented on the basis of an eight channel biopotential measurement device, Wi-Fi module as well as signal processing and classification provided as a Cloud service. In the system, the real-time sEMG data stream is filtered, feature extracted and classified within each data segment and the processed data is visualized in a browser remotely together with the classification result.