Generative Model Based Highly Efficient Semantic Communication Approach for Image Transmission

• DOI: 10.48550/arxiv.2211.10287
• Publication Date: 2022-01-01
• Authors: Tianxiao Han; Jiancheng Tang; Qianqian Yang; Yiping Duan; Zhaoyang Zhang; Zhiguo Shi
• Keywords: Representation; Generative model

Deep learning (DL) based semantic communication methods have been explored to transmit images efficiently in recent years. In this paper, we propose a generative model further improve the efficiency of image transmission and protect private information. particular, transmitter extracts interpretable latent representation from original by exploiting GAN inversion method. We also employ privacy filter knowledge base erase information replace it with natural features base. The simulation results indicate that our proposed method achieves comparable quality received while significantly reducing costs compared existing methods.