A Dynamic Convolution Kernel Generation Method Based on Regularized Pattern for Image Super-Resolution

doi:10.3390/s22114231

Return

A Dynamic Convolution Kernel Generation Method Based on Regularized Pattern for Image Super-Resolution

2022-06-01

Hesen Feng

Lihong Ma *

Jing Tian

DOI：10.3390/s22114231

Original

Save

View PDF

Abstract

En 中文

Image super-resolution aims to reconstruct a high-resolution image from its low-resolution counterparts. Conventional image super-resolution approaches share the same spatial convolution kernel for the whole image in the upscaling modules, which neglect the specificity of content information in different positions of the image. In view of this, this paper proposes a regularized pattern method to represent spatially variant structural features in an image and further exploits a dynamic convolution kernel generation method to match the regularized pattern and improve image reconstruction performance. To be more specific, first, the proposed approach extracts features from low-resolution images using a self-organizing feature mapping network to construct regularized patterns (RP), which describe different contents at different locations. Second, the meta-learning mechanism based on the regularized pattern predicts the weights of the convolution kernels that match the regularized pattern for each different location; therefore, it generates different upscaling functions for images with different content. Extensive experiments are conducted using the benchmark datasets Set5, Set14, B100, Urban100, and Manga109 to demonstrate that the proposed approach outperforms the state-of-the-art super-resolution approaches in terms of both PSNR and SSIM performance.

Keywords:

image super-resolution

dynamic convolution kernel

regularized pattern

multi-task learning

RPB-RDN

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.