Visualization and Analysis of Mental States Based on Photoplethysmogram

In this paper, we analyze fatigue of subjects based on photoplethysmogram (PPG) signal using dimensionality reduction techniques. Various kinds of features can be extracted from PPG, they, however, often represent different meanings. It means that some relationships between PPG signals can be defined based on different measures. Using these characteristics, the set of PPG signal measured in several conditions are mapped into low dimensional space using multiple relational embedding (MRE), and the results can be visually analyzed for finding some important information to estimate fatigue.