cXR+ Voxel-Based Semantic Compression for Networked Immersion

Extended Reality (XR) applications, which may encompass Augmented (AR) and Virtual (VR), commonly involve immersive virtual environments that are based on real physical environments. Transmitting the extensive color depth image data for representing a environment over communication network to create corresponding at remote location is very challenging due enormous volumes time constraints of real-time immersion. We explore semantic compression, conveys meaning through color-codes (CCs) reduce transmitted volume. The creation an conventionally involves five steps (with output): calibration (single-frame point cloud), registration (registered volume reconstruction (voxel frame), marching cubes (meshes), rendering (3D environment). develop novel cXR+ compression splits into client-side function (VNF) represents registered cloud as CCs transmission server-side VNF decodes obtain voxel frames so complete reconstruction. consider analogous splitting two VNFs with Graph Coloring (GC) comparison benchmark. Additionally, we raw JPG compressed well subsequent benchmarks. conduct measurements (computing) time, ratio, accuracy real-world sets elucidate trade-offs in employing voxel-based transmitting networks.