SREFBN: Enhanced feature block network for single‐image super‐resolution

Abstract

Deep learning has assisted the field of single‐image super‐resolution (SR) in achieving new heights. However, the task of restoring a high‐resolution (HR) image from a highly degraded low‐resolution (LR) image is sophisticated due to poor image restoration quality. A novel and effective lightweight SR method is presented as super‐resolution via an enhanced feature block network (SREFBN) that successfully reconstructs an HR image using a corresponding LR image with a purposed deep residual block. In addition, a novel shared parameters approach in the top‐down pathway among low‐level feature maps is introduced. The experimental results prove that SREFBN achieves remarkable performance. The presented framework requires lower computational cost and outperforms many state‐of‐the‐art methods. It is also highly adaptable with low‐end devices, requiring lower multiplication and adding operations. A trade‐off comparison between the number of parameters, execution time, and accuracies is given while also showing different variations of our approach to prove the effectiveness and reliability of the shared parameters. Most importantly, the results indicate that our framework has gained state‐of‐the‐art performance on larger scales 3 and 4. Code is available at https://github.com/curzii23/SREFBN.

A Positive-Unlabeled Generative Adversarial Network for Super-Resolution Image Reconstruction Using a Charbonnier Loss

“…For comparison, the PSNR and SSIM values of all experiments were calculated on the Y channel of the experimental image. Six super-resolution reconstruction methods were contrasted with our method (SRPUGAN-Charb), including Bicubic, SRResNet [32], SRMD [3], GASRMD [15], ELapCGAN [24] and SREFBN [25].…”

“…It can be seen that the a-Charbon regularization constraint enhanced the performance and stability of the reconstructed image. To further showcase the superiority of our SRPUGAN-Charbon, widely controlled experiments were conducted with six state-of-the-arts: Bicubic, SRGAN [32], SRResNet [32], SRMD [3], GASRMD [15], ELapCGAN [24] and SREFBN [25]. The six methods and our method were evaluated on the three benchmark test datasets, using PSNR and SSIM.…”

“…Moreover, our method was qualitatively compared with the six state-of-the-arts on ×4 SR. Figures 4 and 5 report the reconstruction results of Bicubic, SRResNet, SRMD, GASRMD, ELapCGAN [24] and SREFBN [25], as well as those of our SRPUGAN-Charb. It can be learned that our method recovered more details (e.g., textures), and provided more natural perceived quality than the other methods.…”

“…Zhang et al [24] developed an enhanced Laplace pyramid network, and trained the network with multilevel CGAN, VGG and robust Charbonnier loss function, producing high-quality super resolution results. Ketsoi et al [25] improved the feature pyramid network into the enhanced feature block network (EFBN), which adopts high-dimensional feature maps, and uses shared parameters among low-level feature maps. In this way, the computational complexity is reduced, and the perceived quality is improved for the reconstructed images.…”

Single‐image super‐resolution using lightweight transformer‐convolutional neural network hybrid model

“…Ketsoi et al. [14] propose a pioneering and competent lightweight SR model, the enhanced feature block network (SREFBN), by introducing a new bottom‐up path shared parameter approach that can productively regenerate HR images from the correlative LR images.…”