Low-Latency Point Cloud Video Streaming Using User Adaptive Point-Wise Compression

With the advancement of immersive services such as Extended Reality (XR), there is an increasing demand for high-quality and low-latency transmission methods for 3D point clouds. However, existing point cloud transmission methods face challenges such as long latency in the compression process, as well as degradation in visual quality. This study proposes a novel point cloud transmission method that simultaneously achieves high visual quality and low compression latency. Specifically, the proposed method determines the importance of each point in the point cloud based on the user’s viewpoint, and realizes point-wise adaptive compression by loading and discarding multiple 3D points in parallel using multiple processing cores. Experimental results on the 8iVFBv2 point cloud dataset demonstrate that the proposed method achieves a frame rate that is 17.3× higher than existing low-latency and user-viewpoint-based compression techniques, while maintaining comparable visual quality.