SuppleView: decreasing physically limitations on the movement imitation with viewing motions in the video

説明

Development of 3D joints inference technology from 2D RGB video enables us to create 3D annotations for each frame of a video including movements. We here proposed SuppleView which is supple about the user's physically posing while viewing a video and enables the coordinate translation free viewing. By inferring the 3D joints of frames in a video, we create the 3D model as an actor of movements which are same with the actor's one in an original 2D video. The system transitions these two actor depends on the physical rotation of the user's head. The angle of view for observing the actor also changes. Hence rendering contents of this viewer can be ease for user to move one's body to be same with a target one. From the results of the prototype system, we considered our proof of concept had worked properly in the actual scene of movement imitation. We assumed that SuppleView has a potential to ease movement imitations by decreasing physically limitations on viewing motion with traditional displays.

収録刊行物

詳細情報 詳細情報について

問題の指摘

ページトップへ