表情, 心拍, 呼吸数データを用いた人の疲労度推定

抄録

This research aims to develop a system that estimates human fatigue in daily situation as a function of lifestyle support robot. For a comfortable use, the system should estimate fatigue in a short time. To be used for a mobile robot, It is possible to use non-contact sensors. Past reseach used several human vital information like heart rate, brain signals, gait, and facial expression. We use heart rate, facial expression, and respiration rate as a features to estimate fatigue. As an indicator of facial expression we use Action Units (AU) of Facial Action Coding System. We collected these data using a Kinect and a microwave sensor. After the data collection, we verified a certain level of effectiveness of the proposed method by classification experiment using Suppot Vector Machine and the collected data.