.png){kind=icon}
返回
A Multilayer Joint Regularized Block Sparse Bayesian Algorithm With Space-Time Structured Prior Learning Function
.png){kind=icon}
PRE
AI
DOI：10.1109/TGRS.2024.3386732.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
The structured sparse model incorporates dependencies among coefficients, and exploiting intrablock correlations significantly enhances recovery performance. The reconstruction performance of existing block sparse recovery algorithms relies on prior knowledge of blockwise sparse coefficients, including the number of blocks, boundaries, and sizes, which severely restricts the applicability of algorithms in scenarios like radar reconstruction. This article proposes a regularized block sparse Bayesian algorithm with structured prior learning, which involves a two-stage processing procedure. The first stage includes the acquisition of space-temporal hierarchical prior information. It introduces a regularization penalty term in the cost function to constrain and optimize hyperparameters, thereby promoting block correlations among neighboring coefficients. In the second stage, refined recovery is executed for nonzero blocks using a model with multiple measurement vectors (MMVs) constructed from space-temporal dimensional data. This model captures temporal correlations among multiple snapshots and blocks correlations within the prior coefficient structure, enabling joint parameter reconstruction across multiple domains. Simulation results demonstrate that compared with the conventional block sparse Bayesian learning (BSBL) algorithm, the proposed approach exhibits significant advantages in reconstruction performance under an unknown block sparse model.
Keyword:
Block sparse recovery
compressive sensing
frequency agile radar
multidomain joint reconstruction
sparse Bayesian learning (SBL)
期刊
IF:
8.6
论文数:
2.1W
被引数:
10.7W
机构
引用论文
Effect of the C‐terminal proline repeats on ordered packing of squid rhodopsin and its mobility in membranes.png)
FEBS Letters
OAAI
FEBS LettersIF0
OAAI分享收藏Zinc‐Oxynitride TFTs: Toward a New High‐Mobility Low‐Cost Thin‐Film Semiconductor锌氧氮化物TFTs：迈向一种新型高迁移率低成本薄膜半导体
PREAI
分享收藏Pattern-Coupled Sparse Bayesian Learning for Inverse Synthetic Aperture Radar Imaging模式耦合稀疏贝叶斯学习在逆合成孔径雷达成像中的应用
PREAI
分享收藏Joint cognitive optimization of transmit waveform and receive filter against deceptive interference抗欺骗性干扰的发射波形和接收滤波器的联合认知优化SIGNAL PROCESSING
SIGNAL PROCESSINGIF3.6
PREAI
分享收藏Characterizing the clinical relevance of epithelial mesenchymal transition (EMT) in human tumors.
PREAI
分享收藏
查看更多内容