图神经网络与近似消息传递算法

最新推荐文章于 2024-08-14 00:23:17 发布

ZhengXinTang

最新推荐文章于 2024-08-14 00:23:17 发布

阅读量610

点赞数 9

分类专栏： # 图神经网络文章标签：神经网络算法 python

本文链接：https://blog.csdn.net/chumingqian/article/details/137687058

版权

图神经网络专栏收录该内容

22 篇文章 1 订阅

订阅专栏

1. 压缩感知的介绍

https://zhuanlan.zhihu.com/p/22445302

1.1 近似消息传递算法

https://etc.lmdewz.xyz/?p=94
https://blog.csdn.net/qq_53305859/article/details/134355434；

2. 图网络与AMP

有研究工作将图神经网络（GNN）与近似消息传递（AMP）算法和压缩传感相结合。以下是一些相关研究：

Graph Neural Network Enhanced Approximate Message Passing for MIMO Detection:用于 MIMO 检测的图神经网络增强近似消息传递：
- This study proposes a low complexity GNN enhanced AMP algorithm, AMP-GNN, for multiple-input multiple-output (MIMO) detection. The GNN module addresses the inaccuracy of Gaussian approximation for multiuser interference cancellation in AMP detectors. The results show significant performance improvements over traditional AMP detectors.

该文提出了一种低复杂度的GNN增强AMP算法AMP-GNN，用于多输入多输出（MIMO）检测。GNN 模块解决了 AMP 探测器中多用户干扰消除的高斯近似不准确的问题。结果显示，与传统的AMP检测器相比，性能有了显著的提高。

(He et al., 2022).

AMP-Net: Denoising-Based Deep Unfolding for Compressive Image Sensing:
- AMP-Net, a deep unfolding model, is established by unfolding the iterative denoising process of the AMP algorithm. This model integrates deblocking modules to eliminate blocking artifacts in compressive sensing of visual images. The sampling matrix is jointly trained with other network parameters to enhance reconstruction performance.

通过展开AMP算法的迭代去噪过程，建立了AMP-Net深度展开模型。该模型集成了去阻塞模块，以消除视觉图像压缩传感中的阻塞伪影。采样矩阵与其他网络参数联合训练，增强重建性能。

(Zhang et al., 2020).

MB-DAMPNet: A Novel Multi-Branch Denoising-Based Approximate Message Passing Algorithm via Deep Neural Network for Image Reconstruction:MB-DAMPNet：

一种基于深度神经网络的基于多分支去噪的近似消息传递算法，用于图像重建：

MB-DAMPNet introduces a multi-branch structure combining sampling, initial recovery, and deep recovery subnets. It optimizes these components jointly to improve the accuracy and flexibility of image reconstruction from compressive sensing measurements.MB-DAMPNet 引入了一种多分支结构，结合了采样、初始恢复和深度恢复子网。它共同优化了这些组件，以提高压缩传感测量图像重建的准确性和灵活性。
(Yue et al., 2021).

These studies demonstrate the potential of integrating GNNs with AMP algorithms and compressive sensing techniques to enhance performance in various applications such as MIMO detection and image reconstruction.这些研究表明，将GNN与AMP算法和压缩传感技术相结合，可以提高MIMO检测和图像重建等各种应用的性能。