Abstract
While Virtual Reality (VR) retains industry interest, the overall market adoption of VR is still slow. One problem in VR is the social isolation. Many applications are still single user-driven and VR services that do allow the sharing of virtual experiences with others, mostly focus on representing users as artificial avatars. This might be good in some cases (e.g. gaming) but may be too restrictive for interactions where non-verbal communication is important, such as video conferencing, watching videos together, and remote collaboration. For such cases, the importance of photo-realistic VR is acknowledged within the industry and standardisation bodies have begun to address the technology gaps in order to create such services. As an example, MPEG included a generic Social VR architecture into its recent OMAF specification. Also, 3GPP recently started new work on 360-degree VR conferencing and is studying the relevance of many more VR (and AR) communication use cases and technologies to enable them in 5G. In this paper, we report on our recent experiments allowing people to interact, communicate and collaborate with each other as if they were in the same place while sharing a virtual environment. Our VR conferencing platform is modular, web-based and allows for rapid creation of photo-realistic shared experiences. We evaluated our platform with 313 users in six experiences, for both 360-degree video and 3D volumetric VR. We describe the conditions of each experience, reflect on participant survey results, and sketch a roadmap for the future standardisation of social VR services that address, both shared photo-realistic VR experiences in 360-degree, and volumetric video.
Introduction
The last few years have seen a major uptake of virtual reality technology, enabling the creation of immersive video games and training applications, and also paving the way for new forms of video entertainment. One key challenge that many of those VR experiences face is the social barrier. That is the apparent discrepancy between the physical separation of wearing a head-mounted display (HMD) and the human need for sharing their experiences. This can also be seen by large investments into Social VR from key industry companies such as Facebook, Microsoft and HTC.
However, currently, the efforts of large industry players mainly focus on artificial and avatar-based representations of people for use in communication applications. Even though this is good for some use cases, avatar-based approaches may be too restrictive for interactions where non-verbal communication is important, such as video conferencing, social gatherings, presentations, watching 360-degree videos together, intense remote collaboration and many more. Particularly, when it comes to talking to people like family and friends, it is important to see a natural representation of them.
To address this problem, we developed a VR framework that extends current video conferencing capabilities with new VR functionalities. Our framework is modular, based on web technologies and allows both, the easy creation of VR experiences that are social, and the consumption of them with off-the-shelf hardware. With our framework, we aim to allow users to interact or collaborate while being immersed in interactive VR content.
In this paper, we present six social VR experiences which we have built with the help of our framework. We evaluated all six experiences with a total of 313 people, in unstructured testing sessions having a duration of 1-10 minutes. Each testing session was followed by a questionnaire and an informal discussion with a focus on assessing overall quality, video quality, audio quality, presence and immersion in photo-realistic social VR.
No comments yet