Lightweight image super-resolution with enhanced CNN

Abstract

Deep convolutional neural networks (CNNs) with strong expressive ability have achieved impressive performances on single image super-resolution (SISR). However, their excessive amounts of convolutions and parameters usually consume high computational cost and more memory storage for training a SR model, which limits their applications to SR with resource-constrained devices in real world. To resolve these problems, we propose a lightweight enhanced SR CNN (LESR-CNN) with three successive sub-blocks, an information extraction and enhancement block (IEEB), a reconstruction block (RB) and an information refinement block (IRB). Specifically, the IEEB extracts hierarchical low-resolution (LR) features and aggregates the obtained features step-by-step to increase the memory ability of the shallow layers on deep layers for SISR. To remove redundant information obtained, a heterogeneous architecture is adopted in the IEEB. After that, the RB converts low-frequency features into high-frequency features by fusing global and local features, which is complementary with the IEEB in tackling the long-term dependency problem. Finally, the IRB uses coarse high-frequency features from the RB to learn more accurate SR features and construct a SR image. The proposed LESRCNN can obtain a high-quality image by a model for different scales. Extensive experiments demonstrate that the proposed LESRCNN outperforms state-of-the-arts on SISR in terms of qualitative and quantitative evaluation. The code of LESR-CNN is accessible on https://github.com/hellloxiaotian/LESRCNN.

DRNet: A Deep Neural Network With Multi-Layer Residual Blocks Improves Image Denoising

“…A deep structured network was proposed for compressive sensing reconstruction of natural images, combined with the structural insight of traditional optimization methods and the superior speed [30]. To make the deep convolution neural network can be applied to train SR model on real devices with less dataset, a lightweight enhanced SRCNN is a workable methods [31]. This method is also used in image denoising.…”

Multi-scale Xception based depthwise separable convolution for single image super-resolution

“…Usually, used high computational cost and more memory consumption for training a SR model. To resolve these problems Tian et al [ 69 ] proposed the lightweight enhanced super-resolution based SRCNN known as (LESRCNN). In this approach authors are used the three types of successive blocks as an information extraction, enhancement, and reconstruction block with information refinement block.…”

Residual Attention Block Search for Lightweight Image Super-Resolution

“…By introducing the design of residual dense block (RDB) [2], we assume the convolution number is 2 and the growth rate is 16. The basic convolution is replaced with residual block, shallow residual block [17] and lightweight residual block [18], respectively. Consequently, we get 5 types of candidate operations while adding skip connection and none [10], which correspond to OP 1 ∼ OP 5 as depicted in Fig.…”