TY - JOUR
T1 - Towards Immersive Metaverse Experience
T2 - A Wireless Adaptive 3D Human Modeling System
AU - Yin, Mingrui
AU - Sen, Sohom
AU - Guan, Yongjie
AU - Hou, Xueyu
AU - Han, Tao
AU - Ansari, Nirwan
N1 - Publisher Copyright:
© 1986-2012 IEEE.
PY - 2025
Y1 - 2025
N2 - The concept of the metaverse envisions an immersive environment where people can interact, participate in 3D meetings, engage in immersive gaming, and work, and experience the seamless integration of real and virtual worlds. Mixed reality (MR) systems hold significant potential to realize this vision. However, current MR environments face key technical challenges, such as limited network bandwidth and computing capacity. This paper addresses these issues by proposing A wireless adaptive 3D human modeling system calledWadmanMR. WadmanMR dynamically adapts to varying computational and network conditions through a sophisticated integration of smart cameras, servers, and MR headsets, enabling real-time 3D human modeling from 2D images and videos. Performance evaluations of WadmanMR demonstrate that it can maintain real-time processing standards, achieving an average end-to-end latency of less than 100ms and packet loss below 1%, under varying network conditions. WadmanMR outperforms the current state-of-the-art like MagicStream, ExPose and Pixie. Furthermore, this paper explores several open research issues aimed at further improving the performance and applicability of MR systems.
AB - The concept of the metaverse envisions an immersive environment where people can interact, participate in 3D meetings, engage in immersive gaming, and work, and experience the seamless integration of real and virtual worlds. Mixed reality (MR) systems hold significant potential to realize this vision. However, current MR environments face key technical challenges, such as limited network bandwidth and computing capacity. This paper addresses these issues by proposing A wireless adaptive 3D human modeling system calledWadmanMR. WadmanMR dynamically adapts to varying computational and network conditions through a sophisticated integration of smart cameras, servers, and MR headsets, enabling real-time 3D human modeling from 2D images and videos. Performance evaluations of WadmanMR demonstrate that it can maintain real-time processing standards, achieving an average end-to-end latency of less than 100ms and packet loss below 1%, under varying network conditions. WadmanMR outperforms the current state-of-the-art like MagicStream, ExPose and Pixie. Furthermore, this paper explores several open research issues aimed at further improving the performance and applicability of MR systems.
UR - http://www.scopus.com/inward/record.url?scp=105006829050&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=105006829050&partnerID=8YFLogxK
U2 - 10.1109/MNET.2025.3573906
DO - 10.1109/MNET.2025.3573906
M3 - Article
AN - SCOPUS:105006829050
SN - 0890-8044
JO - IEEE Network
JF - IEEE Network
ER -