This paper presents an end-to-end multi-scale deep encoder (convolution) and decoder (deconvolution) network for single image super-resolution (SISR) guided by phase congruency (PC) edge map. Our system starts by a single scale symmetrical encoder-decoder structure for SISR, which is extended to a multi-scale model by integrating wavelet multi-resolution analysis into our network. The new multi-scale deep learning system allows the low resolution (LR) input and its PC edge map to be combined so as to precisely predict the multi-scale super-resolved edge details with the guidance of the highresolution (HR) PC edge map. In this way, the proposed deep model takes both the reconstruction of image pixels' intensities and the recovery of multiscale edge details into consideration under the same framework. We evaluate the proposed model on benchmark datasets of different data scenarios, such as Set14 and BSD100 -natural images, Middlebury and New Tsukubadepth images. The evaluations based on both PSNR and visual perception reveal that the proposed model is superior to the state-of-the-art methods.
Face hallucination aims to produce a high‐resolution face image from an input low‐resolution face image, which is of great importance for many practical face applications, such as face recognition and face verification. Since the structure of the face image is complex and sensitive, obtaining a super‐resolved face image is more difficult than generic image super‐resolution. Recently, with great success in the high‐level face recognition task, deep learning methods, especially generative adversarial networks (GANs), have also been applied to the low‐level vision task – face hallucination. This work is to provide a model evolvement survey on GAN‐based face hallucination. The principles of image resolution degradation and GAN‐based learning are presented firstly. Then, a comprehensive review of the state‐of‐art GAN‐based face hallucination methods is provided. Finally, the comparisons of these GAN‐based face hallucination methods and the discussions of the related issues for future research direction are also provided.
Feature fusion is an important part of building high-precision convolutional neural networks. In the field of image classification, though widely used in processing multiscale features of the same layer and short connections in the same receptive field, feature fusion is rarely used in long connection operations across receptive fields. In order to fuse the high- and low-level features of image classification, a feature fusion module SCFF (selective cross-layer feature fusion) for long connections is designed in this work. The SCFF can connect the long-distance feature maps in different receptive fields in a top-down order and apply the self-attention mechanism to fuse them two by two. The final fusion result is used as the input of the classifier. In order to verify the effectiveness of the model, the image classification experiment was done on a number of typical datasets. The experimental results prove that the model can fit the existing convolutional neural network well and effectively improve the classification accuracy of the convolutional network only at the cost of a small amount of calculation.
Resolution decrease and motion blur are two typical image degradation processes that are usually addressed by deep networks, specifically convolutional neural networks (CNNs). However, since real images are usually obtained through multiple degradations, the vast majority of current CNN methods that employ a single degradation process inevitably need to be improved to account for multiple degradation effects. In this work, motivated by degradation decoupling and multiple-order attention drop-out gating, we propose a joint deep recovery model to efficiently address motion blur and resolution reduction simultaneously. Our degradation decoupling style improves the continence and the efficiency of model construction and training. Moreover, the proposed multi-order attention mechanism comprehensively and hierarchically extracts multiple attention features and fuses them properly by drop-out gating. The proposed approach is evaluated using diverse benchmark datasets including natural and synthetic images. The experimental results show that our proposed method can efficiently complete joint motion blur and image super-resolution (SR).
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.