Comparative Study of AR Versus Image and Video for Exercise Learning

Abstract: There is inadequate attention to using mobile Augmented Reality (AR) in fitness, despite mobile AR being easy to use, requiring no extra cost, and can be a powerful learning tool. In this work, we present a mobile AR application that can help users learn exercises with a virtual personal trainer. We conduct a user study with 10 participants to investigate the learning quality of the ARFit (i.e., the proposed mobile AR application) in comparison to traditional methods such as Image-based learning and Video-based learning. Our results indicate that participants have a higher learning quality of exercise with mobile AR than (1) Image-based learning among all exercises selected and (2) video-based learning with exercise that requires greater spatial knowledge, with the performance evaluated by a qualified personal trainer. In addition, ARFit has an excellent rating in usability, is deemed to be highly acceptable, and is the preferred exercise learning method by most participants (N=9)