Supporting Immersive Viewer Engagement with Virtual YouTubers

Abstract

Live streaming has gained more popularity in the digital world, and VR technologies have been widely engaged in live streaming for entertainment. Nonetheless, the chatroom is the only approach for viewers to communicate with streamers in traditional live streaming and is limited to input hardware.

Meanwhile, Virtual YouTubers (VTubers) represent a virtual and digital culture trend and are growing rapidly, but their interactions with viewers are similar to real-person streamers, lacking innovation and technological sense. Viewers merely use flat user interfaces of 2D video platforms and send comments through keyboards. Though previous works tried to engage VR technologies in VTubers' live streams, the users' interactions are still limited to self movements and simple functions. The overloaded chat comments and traditional text input methods are also the fundamental problems limiting the interactions and user experience.

This project proposes a new combination of VTuber live concerts with VR environments and supports more convenient input methods to achieve pre-set and customised interactions among VTubers and viewers, further improving the user experience. The possible interactions include speech recognition, virtual keyboard input, general movement synthesization, music beats-based cheer-up movement synthesization, and voting rights. Other features are also designed to increase users' sense of presence, such as Danmaku and lyrics displayed in the virtual environment.