[2D1] Haptics Display - Multimodal

[2D1-05] Appending Vibrotactile Sensation for First-Person Videos using Machine Learning

○Kentaro Yoshida¹, Yuuki Horiuchi¹, Yasutoshi Makino¹, Hiroyuki Shinoda¹ (1. The University of Tokyo)

Recent advances in video technology are remarkable. Especially with the spread of VR technology and action cameras, first-person video become widely popular. In addition, it has been reported that quality of experience improves by adding tactile sensation to audiovisual-only videos. In this paper, we propose a method to estimate suitable vibrotactile sensation to videos composed only of sounds and sequential images, based on a video recorded by the action camera and the acceleration data recorded at the same time. We also propose a method to present the tactile sensation.