Task‐driven latent active correction for physics‐inspired input method in near‐field mixed reality applications

• DOI: 10.1002/jsid.728
• Publication Date: 2018-08-13T00:19:23Z
• Authors: Zhenliang Zhang; Yue Li; Jie Guo; Dongdong Weng; Yue Liu; Yongtian Wang
• Keywords: 05 social sciences; 0501 psychology and cognitive sciences

Abstract Calibration accuracy is one of the most important factors to affect user experience in mixed reality applications. For a typical system built with optical see‐through head‐mounted display, key problem how guarantee hand–eye coordination by decreasing instability eye and display long‐term use. In this paper, we propose real‐time latent active correction algorithm decrease calibration errors accumulated over time. Experimental results show that can an effective result improve proposed algorithm. Based on system, experiments about virtual buttons are also designed, interactive performance regarding different scales presented. Finally, direct physics‐inspired input method constructed, which shares similar gesture‐based but provides lower learning cost due its naturalness.