广东工业大学学报 ›› 2024, Vol. 41 ›› Issue (04): 114-121.doi: 10.12052/gdutxb.230103

• 计算机科学与技术 • 上一篇    下一篇

基于深度学习的自适应采样及记忆增强压缩感知方法

罗成, 张军   

  1. 广东工业大学 信息工程学院 , 广东 广州 510006
  • 收稿日期:2023-08-04 出版日期:2024-07-25 发布日期:2024-08-13
  • 通信作者: 张军(1979–) ，男，教授，博士，主要研究方向为人工智能技术、压缩感知理论及其应用，E-mail：jzhang@gdut.edu.cn
  • 作者简介:罗成(1998–) ，男，硕士研究生，主要研究方向为压缩感知、深度学习等，E-mail：lc20162020@163.com
  • 基金资助:
    国家自然科学基金资助项目(61973088)

Adaptive Sampling and Memory-augmented Compressed Sensing Algorithm Based on Deep Learning

Luo Cheng, Zhang Jun   

  1. School of Information Engineering, Guangdong University of Technology, Guangzhou 510006, China
  • Received:2023-08-04 Online:2024-07-25 Published:2024-08-13

HTML

PDF (PC)

26

摘要： 深度学习技术在压缩感知重构中的应用显著提高了重构的速度和精度。然而现有的深度压缩感知算法通常采用相同的采样率来处理不同的块，忽视了不同图像块具有不同重构难度的事实。本文提出了一种自适应采样与记忆增强的压缩感知算法。首先，本文基于测量域的重构误差估计不同块的重构难度，然后设计规则来自适应分配采样率，采样矩阵则通过掩码实现在特定采样率下的图像块采样。进一步地，在重构网络中增加双支路融合模块来增强上下文记忆的交互，通过调整不同记忆支路的通道权重，提高了网络的重构能力。实验结果表明：与其他方法相比，所提出的算法在几个常用数据集上的平均结构相似性(Structural Similarity, SSIM) 提高了0.0269，平均峰值信噪比(Peak Signal-to-Noise Ratio, PSNR) 提高了1.66 dB。

关键词: 深度学习, 自适应采样, 记忆增强

Abstract: The deep learning technology has significantly improved the speed and accuracy of compressed sensing reconstruction. However, the existing deep compressive sensing algorithms usually use the same sampling rate to process different blocks of an image, ignoring the fact that different image blocks have different reconstruction difficulties. In this paper, a compressive sensing algorithm with adaptive sampling and memory enhancement is proposed. Firstly, the reconstruction difficulty of different blocks is estimated based on the reconstruction error of the measurement domain. Then, the rules are designed to adaptively assign the sampling rates, and the sampling matrix is used to sample each image block at a specific sampling rate with the help of a sampling rate mask. Furthermore, the two-branch aggregation module is added to the reconstruction network to enhance the interaction of context memory, and the reconstruction ability of the network is improved by adjusting the channel weight of different memory branches. The experimental results show that the proposed algorithm increases the average SSIM by approximately 0.0269 and the average PSNR by approximately 1.66 dB over other methods on several common datasets.

Key words: deep learning, adaptive sampling, memory-augmented

中图分类号: 

  • TP391
[1] DONOHO D L. Compressed sensing [J]. IEEE Transactions on Information Theory, 2006, 52(4): 1289-1306.
[2] CANDES E J, WAKIN M B. An introduction to compressive sampling [J]. IEEE Signal Processing Magazine, 2008, 25(2): 21-30.
[3] LUSTIG M, DONOHO D L, SANTOS J M, et al. Compressed sensing MRI [J]. IEEE Signal Processing Magazine, 2008, 25(2): 72-82.
[4] ENDER J H G. On compressive sensing applied to radar [J]. Signal Processing, 2010, 90(5): 1402-1414.
[5] ZHENG S, CHEN J, ZHANG X P, et al. A new multihypothesis-based compressed video sensing reconstruction system [J]. IEEE Transactions on Multimedia, 2021, 23: 3577-3589.
[6] KULKARNI K, LOHIT S, TURAGA P, et al. Reconnet: non-iterative reconstruction of images from compressively sensed measurements[C]// IEEE Conference on Computer Vision & Pattern Recognition (CVPR) . Las Vegas: IEEE, 2016: 449-458.
[7] SHI W Z, JIANG F, LIU S H, et al. Image compressed sensing using convolutional neural network [J]. IEEE Transactions on Image Processing, 2019, 29: 375-388.
[8] SHI W Z, JIANG F, LIU S H, et al. Scalable convolutional neural network for image compressed sensing[C]// IEEE Conference on Computer Vision and Pattern Recognition (CVPR) . Long Beach: IEEE, 2019: 12282-12291.
[9] BLUMENSATH T, DAVIES M E. Iterative hard thresholding for compressed sensing [J]. Applied and Computational Harmonic Analysis, 2009, 27(3): 265-274.
[10] ZHANG J, GHANEM B. Istanet: interpretable optimization-inspired deep network for image compressive sensing[C]// IEEE Conference on Computer Vision and Pattern Recognition (CVPR) . SaltLake City: IEEE, 2018: 1828-1837.
[11] YOU D, XIE J F, ZHANG J. Ista-net++: flexible deep unfoldingnetwork for compressive sensing[C]// IEEE International Conference on Multimedia and Expo (ICME) . Shenzhen: IEEE, 2021: 1-6.
[12] YU Y, WANG B, ZHANG L M. Saliency-based compressive sampling for image signals [J]. IEEE Signal Processing Letters, 2010, 17(11): 973-976.
[13] ZHOU S W, HE Y, LIU Y H, et al. Multi-channel deep networksfor block-based image compressive sensing [J]. IEEE Transactionson Multimedia, 2021, 23: 2627-2640.
[14] CANDES E J. The restricted isometry property and its implications for compressed sensing [J]. Comptes Rendus Mathematique, 2008, 346: 589-592.
[15] QIU C X, YUE T, HU X M. Adaptive and cascaded compressivesensing[EB/OL]. arXiv: 2203.10779 (2022-3-21) [2023-8-3]. https://doi.org/10.48550/arXiv.2203.10779.
[16] DONOHO D L, MALEKI A, MONTANARI A. Message-passing algorithms for compressed sensing [J]. Proceedings of the National Academy of Sciences of the United States of America, 2009, 106(45): 18914-18919.
[17] LI C, YIN W, JIANG H, et al. An efficient augmented lagrangian method with applications to total variation minimization [J]. Computational Optimization and Applications, 2013, 56(3): 507-530.
[18] ZHANG J, ZHAO C, GAO W. Optimization-inspired compact deep compressive sensing [J]. IEEE Journal of Selected Topics in Signal Processing, 2020, 14(4): 765-774.
[19] ZHANG Z H, LIU Y P, LIU J, et al. Ampnet: denoising-based deep unfolding for compressive image sensing [J]. IEEE Transactions on Image Processing, 2021, 30: 1487-1500.
[20] YOU D, ZHANG J, XIE J F, et al. Coast: controllable arbitrary sampling network for compressive sensing [J]. IEEE Transactions on Image Processing, 2021, 30: 6066-6080.
[21] CHEN W J, YANG C L, YANG X. Fsoinet: feature-space optimi-zation-inspired network for image compressive sensing[C]// IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). Singapore: IEEE, 2022: 2460-2464.
[22] RONNEBERGER O, FISCHER P, BROX T. U-net: convolutionalnetworks for biomedical image segmentation[C]// International Conference on Medical Image Computing and Computer Assisted Intervention. Munich: Springer, 2015: 234-241.
[23] LI X, WANG W, HU X, et al. Selective kernel networks[C]// IEEE Conference on Computer Vision & Pattern Recognition. LongBeach: IEEE, 2019: 510-519.
[1] 梁宇辰, 蔡念, 欧阳文生, 谢依颖, 王平. 基于切片关联信息的慢性阻塞性肺疾病CT诊断[J]. 广东工业大学学报, 2024, 41(01): 27-33.
[2] 吴菊华, 李俊锋, 陶雷. 基于知识图谱嵌入与深度学习的药物不良反应预测[J]. 广东工业大学学报, 2024, 41(01): 19-26,40.
[3] 温雯, 刘莹, 蔡瑞初, 郝志峰. 面向多粒度交通流预测的时空深度回归模型[J]. 广东工业大学学报, 2023, 40(04): 1-8.
[4] 钟耿君, 李东. 基于通道分离机制的双分支点云处理网络[J]. 广东工业大学学报, 2023, 40(04): 18-23.
[5] 金宇凯, 李志生, 欧耀春, 张华刚, 曾江毅, 陈搏超. 基于多阶段聚类的PM2.5质量浓度预测及对比研究[J]. 广东工业大学学报, 2023, 40(03): 17-24.
[6] 吴俊贤, 何元烈. 基于通道注意力的自监督深度估计方法[J]. 广东工业大学学报, 2023, 40(02): 22-29.
[7] 刘冬宁, 王子奇, 曾艳姣, 文福燕, 王洋. 基于复合编码特征LSTM的基因甲基化位点预测方法[J]. 广东工业大学学报, 2023, 40(01): 1-9.
[8] 徐伟锋, 蔡述庭, 熊晓明. 基于深度特征的单目视觉惯导里程计[J]. 广东工业大学学报, 2023, 40(01): 56-60,76.
[9] 刘洪伟, 林伟振, 温展明, 陈燕君, 易闽琦. 基于MABM的消费者情感倾向识别模型——以电影评论为例[J]. 广东工业大学学报, 2022, 39(06): 1-9.
[10] 章云, 王晓东. 基于受限样本的深度学习综述与思考[J]. 广东工业大学学报, 2022, 39(05): 1-8.
[11] 郑佳碧, 杨振国, 刘文印. 基于细粒度混杂平衡的营销效果评估方法[J]. 广东工业大学学报, 2022, 39(02): 55-61.
[12] Gary Yen, 栗波, 谢胜利. 地球流体动力学模型恢复的长短期记忆网络渐进优化方法[J]. 广东工业大学学报, 2021, 38(06): 1-8.
[13] 赖峻, 刘震宇, 刘圣海. 基于全局数据混洗的小样本数据预测方法[J]. 广东工业大学学报, 2021, 38(03): 17-21.
[14] 岑仕杰, 何元烈, 陈小聪. 结合注意力与无监督深度学习的单目深度估计[J]. 广东工业大学学报, 2020, 37(04): 35-41.
[15] 曾碧, 任万灵, 陈云华. 基于CycleGAN的非配对人脸图片光照归一化方法[J]. 广东工业大学学报, 2018, 35(05): 11-19.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 2017 《广东工业大学学报》编辑部
地址：广州市东风东路729号广东工业大学 邮编：510090
电话：020-37626653,020-37626139,020-37626396
邮箱：xbzrb@gdut.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发