Neural Network-Based Refinement on Multiple Video Frames in V-PCC
Keiichiro Takada, Sujun Hong, Tomohiro Ikai, Yasuaki Tokumo, Takeshi Chujoh, Kakeru Sugimoto · 2025
Video-based point cloud compression (V-PCC) utilizes video compression technology to efficiently encode dense point clouds, achieving state-of-the-art compression performance with a low computational burden. V-PCC converts a 3D point cloud into three sets of 2D frames: occupancy, geometry and attribute, and signals them via the video codec such as HEVC or VVC.