Multi-View Video Semantic Communication

In the 6G era, the demand for higher-quality wireless video transmission with lower communication resources is increasing. For example, multi-view video transmission faces significant technical challenges due to high bandwidth consumption and complex channel environments. To address these issues, this paper proposes an innovative Multi-View Video Semantic Communication (MVSC) framework. By employing a hierarchical semantic encoding mechanism, MVSC leverages the commonality among multi-view videos and integrates joint source-channel coding techniques to achieve efficient data compression and robust transmission. Specifically, the system utilizes individual encoders to extract single-frame features and a spatio-temporal commonality encoder to fuse shared spatiotemporal features across multiple views, substantially reducing information redundancy. Experimental results demonstrate that, under AWGN channels, MVSC outperforms conventional video communication methods in terms of PSNR and MS-SSIM metrics, particularly in bandwidth-constrained and low SNR scenarios. This study provides an efficient and robust solution for multi-view video transmission.