Host: The Japan Society of Mechanical Engineers
Name : [in Japanese]
Date : March 06, 2023 - March 07, 2023
Japan is a super-aging society and the number of people requiring nursing care is increasing with each day. Maintaining the ability to walk by preventing falls and maintaining and improving muscle strength is key to preventing the need for nursing care. However, most of the conventional gait training is monotonous and it is difficult to maintain motivation. Therefore, a system was developed in this study that enables users to enjoy walking and also maintain their motivation. Using CycleGAN, a machine learning model capable of transforming one image to another, a system was developed that enables walking while transforming real scenery in real-time and presenting it on a head-mounted display, and verifies the psychological impact of the proposed system on pedestrians. A 10 m walking experiment was conducted on males in their 20s, and the results obtained were compared using the two-dimensional mood scale-short term (TDMS-ST) method and a subjective evaluation questionnaire using the visual analog scale as a walking evaluation questionnaire. The results of the experiment showed that the arousal level of the TDMS-ST increased and the values of the items "Did you find walking enjoyable?" and "Did you want to continue walking?" in the walking evaluation questionnaire increased. This suggests that the proposed system can assist with the motivation for walking. However, it has also been suggested that it might increase the feeling of fatigue from walking.
