The Electronics and Telecommunications Research Institute (ETRI) announced on the 19th that it has developed an ‘Immersive Extended Reality (XR)’ technology that combines virtual reality (VR) and augmented reality (AR).
The core of immersive XR is the precise three-dimensional motion analysis for multiple users and user position estimation technology in complex spaces. The multi-user three-dimensional precision motion analysis is achieved by using deep learning technology in mobile environments to infer the user’s three-dimensional joint position information based on two-dimensional video data.
Previously, systems equipped with expensive graphics processing units (GPUs) were required, but ETRI has made it possible to efficiently extract three-dimensional posture information using a user’s smartphone.
Additionally, the user position estimation in complex spaces is a technology aimed at seamlessly integrating augmented reality content into the real world. This technology extracts features from video with a smartphone camera, generates and corrects three-dimensional point clouds to build an environmental map, allowing precise estimation of the user’s device location.
A notable advantage of immersive XR is its design that maximizes user immersion in AR environments. This allows for precise interpretation of the user’s environment and behavior in VR and AR settings, enabling the provision of consistent interaction quality to multiple users without any restrictions, according to ETRI.
ETRI also developed trial content to create a virtual space tailored to augmented reality, allowing numerous users to participate in XR experiences both locally and remotely.
In augmented reality, two local users wearing head-mounted displays (HMDs) playing the roles of a fox and the Little Prince can interact with another remote user wearing an HMD playing the role of a scarecrow in virtual reality. Through this, ETRI was able to demonstrate remote immersive extended reality with multiple participants.
This research was jointly conducted over four years by ETRI, KAIST, J2Y Soft, and TOZ, with support from the Ministry of Science and ICT’s projects.