Deep neural networks (DNNs) have been widely adopted in single image super-resolution (SISR) recently with great success. As a network goes deeper, intermediate features become hierarchical. However, most SISR methods based on DNNs do not make full use of the hierarchical features. The features cannot be read directly by the subsequent layers, therefore, the previous hierarchical information has little influence on the subsequent layer output, and the performance is relatively poor. To address this issue, a novel global dense feature fusion convolutional network (DFFNet) is proposed, which can take full advantage of global intermediate features. Especially, a feature fusion block (FFblock) is introduced as the basic module. Each block can directly read raw global features from previous ones and then learns the feature spatial correlation and channel correlation between features in a holistic way, leading to a continuous global information memory mechanism. Experiments on the benchmark tests show that the proposed method DFFNet achieves favorable performance against the state-of-art methods.

The performance of image super-resolution (SR) have been greatly improved with deep convolution neural network (CNN). Despite image SR targets at recovering high-frequency details, most SR methods still focus on generating high-level features via a deep and wide network. They lack the discriminative ability of high-frequency information hidden in the abundant CNN features, thus hindering CNNs to yield better SR results. To tackle this issue, we propose two new attention mechanism: context weighted channel attention (CWCA) and persistent spatial attention (PSA). They modulate abundant features by suppressing the useless features and enhancing the interested ones in a channel-and-spatial manner. The network is then enabled to concentrate more on informative features closely related to the high-frequency components of an image. Furthermore, we propose enhanced attention residual groups with dense connection (EARG-D) to capture not only short-term information but also long-term information to maintain more useful features. Finally, we construct a deep enhanced context attention super resolution network (EASR) for better image reconstruction. Quantitative and qualitative experiments well demonstrate that our proposed method performs better than existing state-of-the-art SR methods.

In recent years, deep-learning-based super-resolution (SR) methods have obtained impressive performance gains on synthetic clean datasets, but they fail to perform well in real-world scenarios due to insufficient real-world training data. To tackle this issue, we propose a conditional-normalizing-flow-based method named IDFlow for image degradation modeling that aims to generate various degraded low-resolution (LR) images for real-world SR model training. IDFlow takes image degradation modeling as a problem of learning a conditional probability distribution of LR images given the high-resolution (HR) ones, and learns the distribution from existing real-world SR datasets. It first decomposes the image degradation modeling into blur degradation modeling and real-world noise modeling. It then utilizes two multi-scale invertible networks to model these two steps, respectively. Before applied into real-world SR, IDFlow is first trained supervisedly on two real-world datasets RealSR and DPED with negative log-likelihood (NLL) loss. It is then used to generate a large number of HR-LR image pairs from an arbitrary HR image dataset for SR model training. Extensive experiments on IDFlow with RealSR and DPED are conducted, including evaluations on image degradation stochasticity, degradation modeling, and real-world super resolution. Two known SR models are trained with IDFlow and named as IDFlow-SR and IDFlow-GAN. Testing results on the RealSR and DPED testing dataset show that not only can IDFlow generate realistic degraded images close to real-world images, but it is also beneficial to real-world SR performance improvement. IDFlow-SR achieves 4× SR performance gains of 0.91 dB and 0.161 in terms of image quality assessment metrics PSNR and LPIPS. Moreover, IDFlow-GAN can super-resolve real-world images in the DPED testing dataset with richer textures and maintain clearer patterns without visible noises when compared with state-of-the-art SR methods. Quantitative and qualitative experimental results well demonstrate the effectiveness of the proposed IDFlow.