Ubiquitous Semantic Metaverse
Ubiquitous semantic metaverses aim to create immersive, interconnected virtual worlds seamlessly integrated with the physical environment through advanced semantic understanding. Current research focuses on efficient resource allocation using federated learning and reinforcement learning algorithms, often incorporating blockchain for security and hyperdimensional computing to reduce computational demands. This field is significant for its potential to revolutionize various sectors, including healthcare, education, and industrial automation, by enabling realistic and interactive digital twins and virtual experiences.
Papers
Resource Allocation of Federated Learning for the Metaverse with Mobile Augmented Reality
Xinyu Zhou, Chang Liu, Jun Zhao
XRBench: An Extended Reality (XR) Machine Learning Benchmark Suite for the Metaverse
Hyoukjun Kwon, Krishnakumar Nair, Jamin Seo, Jason Yik, Debabrata Mohapatra, Dongyuan Zhan, Jinook Song, Peter Capak, Peizhao Zhang, Peter Vajda, Colby Banbury, Mark Mazumder, Liangzhen Lai, Ashish Sirasao, Tushar Krishna, Harshit Khaitan, Vikas Chandra, Vijay Janapa Reddi