High-efficiency feature representation based lightweight reconstruction neural network for image compressive sensing

Qiuyan Liu; Chunling Yang

doi:10.1117/12.3031412

19 July 2024 High-efficiency feature representation based lightweight reconstruction neural network for image compressive sensing

Qiuyan Liu, Chunling Yang

Author Affiliations +

Proceedings Volume 13181, Third International Conference on Electronic Information Engineering, Big Data, and Computer Technology (EIBDCT 2024); 131817X (2024) https://doi.org/10.1117/12.3031412
Event: Third International Conference on Electronic Information Engineering, Big Data, and Computer Technology (EIBDCT 2024), 2024, Beijing, China

Abstract

Compressive sensing reconstruction has a great significance for limited sampling resources. The reconstruction methods combining iterative optimization with deep learning gain favorable development. However, many existing methods are inefficient in the feature representation due to the equal computating resource distribution between reconstruction phases and have a problem of a large number of parameters. In this paper, we propose a new scheme of expanding the reconstruction phase’s subnet width with the depth increasing and design a high-efficiency feature representation based lightweight reconstruction neural network (HEFRNet) for image compressive sensing. Firstly, we introduce an incremental feature map generation subnet with the processing of intrinsic feature maps in inter-phase. Secondly, We design a multidimensional optimization subnet in intra-phase to update and optimize the feature maps. The experiments indicate that HEFRNet has significant advantages in parameters and the effect of image reconstruction.

(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Qiuyan Liu and Chunling Yang "High-efficiency feature representation based lightweight reconstruction neural network for image compressive sensing", Proc. SPIE 13181, Third International Conference on Electronic Information Engineering, Big Data, and Computer Technology (EIBDCT 2024), 131817X (19 July 2024); https://doi.org/10.1117/12.3031412

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available