Our group studies human-computer interaction (HCI) with data science, signal processing, pattern recognition, machine learning approaches. Primary goal of our group is to develop an efficient and convenient HCI interface to make more and more people utilize and enrich our daily lives.
Effect of Daily Stress on Bio and Brain Signals
We investigate the effect of daily stress through the values and parameters of HRV and fNIRS. The results of this study demonstrate the feasibility of an HRV- or fNIRS-based mental stress management system and can be used to design a robust stress management system for future personal mental health assessment.
Bio-signal monitoring using remote photoplethysmogram (rPPG) from facial video
We build deep learning models for heart rate or heart rate variability estimation by using rPPG through infrared (IR) and RGB facial videos. This research covers deep learning based IR and RGB video fusion and deep learning based rPPG estimation.
Mixed Reality-based Human-Animal Interaction for Mental Stress Management
We implement a mixed reality (MR) based human-animal interaction content and experimentally verify its effect of reducing mental stress by bio-signal processing.
Automatic Generation of Mathematical Graph Descriptions for Students with Visual Impairments
We implement a graphical user interface software program that converts mathematical equation into the graph, as well as converts the graph into text and speech description to help visually impaired students learn mathematics.
Predicting tDCS treatment outcomes of patients with PTSD using automated EEG classification
We compare PTSD patient’s EEG before and after tDCS treatment and build deep learning model to predict treatment outcomes based on EEG spectrogram.