Adaptive Densely Connected Super-Resolution Reconstruction

TL;DR

This paper introduces an adaptive dense connection-based super-resolution algorithm that enhances feature utilization and reconstructs high-quality images, outperforming existing methods in PSNR, SSIM, and visual quality.

Contribution

It proposes a novel adaptive dense connection framework with a specialized sub-pixel reconstruction layer for improved single image super-resolution.

Findings

Outperforms state-of-the-art algorithms in PSNR and SSIM.

Enhances high-frequency feature learning through pre-trained SKIP.

Achieves superior visual effects in reconstructed images.

Abstract

For a better performance in single image super-resolution(SISR), we present an image super-resolution algorithm based on adaptive dense connection (ADCSR). The algorithm is divided into two parts: BODY and SKIP. BODY improves the utilization of convolution features through adaptive dense connections. Also, we develop an adaptive sub-pixel reconstruction layer (AFSL) to reconstruct the features of the BODY output. We pre-trained SKIP to make BODY focus on high-frequency feature learning. The comparison of PSNR, SSIM, and visual effects verify the superiority of our method to the state-of-the-art algorithms.