Miyashita, FumiyaAmemiya, TomohiroKitazaki, MichiteruKasamatsu, KeikoYem, VibolIkei, YasushiKakehi, Yasuaki and Hiyama, Atsushi2019-09-112019-09-112019978-3-03868-097-01727-530Xhttps://doi.org/10.2312/egve.20191299https://diglib.eg.org:443/handle/10.2312/egve20191299This paper describes the viewpoint suitable for sports training in virtual reality (VR). We compared first-person and third-person view in the accuracy of cognitive simulation and reproduction of the body part trajectory. From the third-person view, the participants were able to understand 66% of the whole body's movement, and from the first-person view, they were able to understand 52%. However, when observing complex movement such as position grasp of a forearm, the third-person view enabled memorization of the position significantly better than the first-person view. It was suggested that the viewpoint needs to be changed depending on the features of the sports.Humancentered computing → Human computer interaction (HCI) → Interaction paradigms → Virtual reality10.2312/egve.2019129923-24