基于Swin Transformer的联合信源信道编码算法
(四川大学 电子信息学院,四川 成都 610065)
摘要
联合信源信道编码作为语义通信的关键研究方向,已取得初步研究成果。然而,随着图像分辨率的提升,传统的基于卷积神经网络的联合信源信道编码算法在图像语义特征提取方面表现出局限性。为了解决该问题,文章提出了一种基于Swin Transformer的联合信源信道编码算法。该算法首先利用多尺度大核注意力机制初步捕获图像的局部信息和长距离依赖性,然后通过Swin Transformer进一步对图像语义特征进行分层提取和自适应码率编码。实验结果表明,在AWGN和Rayleigh信道模型下,所提出的算法在PSNR和MS-SSIM指标上均优于传统的算法。
关键词
联合信源信道编码;Swin Transformer;多尺度大核注意力
参考文献
- [1] XU J,AI B,CHEN W,et al. Wireless Image Transmission Using Deep Source Channel Coding with Attention Modules [J].IEEE Transactions on Circuits and Systems for Video Technology,2021,32(4):2315-2328.
- [2] 朱晓庆,杨红,陈洪刚,等.面向机器视觉任务的自适应语义通信 [J].通信技术,2023,56(9):1043-1050.
- [3] ZHANG W,ZHANG H,MA H,et al. Predictive and Adaptive Deep Coding for Wireless Image Transmission in Semantic Communication [J].IEEE Transactions on Wireless Communications,2023,22(8):5486-5501.
- [4] YANG K,WANG S,DAI J,et al. SwinJSCC: Taming Swin Transformer for Deep Joint Source-Channel Coding [J/OL].arXiv:2308.09361 [eess.SP].[2024-10-20].https://arxiv.org/abs/2308.09361.
- [5] BOURTSOULATZE E,KURKA D B,GÜNDÜZ D. Deep Joint Source-Channel Coding for Wireless Image Transmission [J].IEEE Transactions on Cognitive Communications and Networking,2019,5(3):567-579.
- [6] YOO H,DAI L,KIM S,et al. On the Role of ViT and CNN in Semantic Communications: Analysis and Prototype Validation [J].IEEE Access,2023,11:71528-71541.
- [7] YE H,LI G Y,HUANG B H F,et al. Channel Agnostic End-to-End Learning Based Communication Systems with Conditional GAN [C]//2018 IEEE Globecom Workshops(GC Wkshps).Abu Dhabi:IEEE,2018:1-5.
- [8] 郭畅,何占豪,杨君刚,等.图像语义通信技术综述与展望 [J].电讯技术,2025,65(2):329-338.
- [9] WANG Y,LI Y S,WANG G,et al. Multi-scale Attention Network for Single Image Super-Resolution [C]//2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).Seattle:IEEE,2024:5950-5960.
- [10] LIU Z,LIN Y T,CAO Y,et al. Swin Transformer: Hierarchical Vision Transformer Using Shifted Windows [C]//2021 IEEE/CVF International Conference on Computer Vision (ICCV).Montreal:IEEE,2021:9992-10002.
作者简介
廖潇(2000.04—),男,汉族,四川仁寿人,硕士研究生在读,研究方向:语义通信。