Search

Existing face super-resolution methods depend on deep convolutional networks (DCN) to recover high-quality reconstructed images. They either acquire information in a single space by designing complex models for direct reconstruction, or employ additional networks to extract multiple prior information to enhance the representation of features. However, existing methods are still challenging to perform well due to the inability to learn complete and uniform representations. To this end, we propose a self-attention learning network (SLNet) for three-stage face super-resolution, which fully explores the interdependence of low- and high-level spaces to achieve compensation of the information used for reconstruction. Firstly, SLNet uses a hierarchical feature learning framework to obtain shallow information in the low-level space. Then, the shallow information with cumulative errors due to DCN is improved under high-resolution (HR) supervision, while bringing an intermediate reconstruction result and a powerful intermediate benchmark. Finally, the improved feature representation is further enhanced in high-level space by a multi-scale context-aware encoder-decoder for facial reconstruction. The features in both spaces are explored progressively from coarse to fine reconstruction information. The experimental results show that SLNet has a competitive performance compared to the state-of-the-art methods., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 Elsevier Ltd. All rights reserved.)

Currently, the design of coal mine roadway support schemes is still mainly based on manual design, engineering analogy, and FLAC model simulation, which has problems such as strong subjectivity, low universality, and insufficient utilization of coal mine support big data. The design method based on expert systems has cumbersome rule setting procedures, large engineering quantities, and low intelligence. Case based reasoning (CBR) and deep learning techniques are introduced into the field of roadway support scheme design. Based on text big data such as coal mine support regulations, support standards, and coal mine roadway geological reports, an intelligent design method for coal mine roadway support scheme is proposed. The method obtains 346 sets of roadway support data from different coal mines, extracts structured data and divides it into input and output parameters, and optimizes the input and output parameters through constant attribute variable filtering and high correlation filtering methods. The method establishes a CBR model and imports the extracted structured data into the CBR model to form a case library of support scheme comparison and selection. The method calculates the similarity between the new roadway support scheme and the historical scheme, and outputs the three historical schemes with the highest similarity for comparison, achieving similar case comparison. BP neural network and long short term memory (LSTM) network are respectively used to establish automatic generation models for coal mine roadway support schemes. By comparing the prediction indicators, it is determined to use the combination of LSTM model and CBR model to establish an intelligent design system for coal mine roadway support scheme. The system is used for the design of auxiliary transportation roadway support scheme in the F6226 working face of Buliangou Coal Mine excavation. Through experiments, it is verified that the deformation of the two sides of the roadway and the maximum displacement of the roof under the system generated scheme are smaller than those under the manual design scheme. The integrity of the roadway roof and two sides is good, the bearing capacity of the surrounding rock is enhanced, and the support effect is significant.

The goal of pan-sharpening tasks is to fuse panchromatic (PAN) images and low-spatial-resolution (LR) multispectral (MS) images for the purpose of aggregating texture and spectral information. Although traditional embedding-based pan-sharpening methods achieve competitive results, they are limited by the shallow network and not suitable for large-scale datasets. In this study, we design a novel multiscale locally linear embedding residual network (LLERN) that consists of two phases: the spectral preservation phase and the structural preservation phase. As the pretreatment of the structural preservation network, the spectral preservation network aims to upscale the LR MS image while retaining spectral information. The proposed locally linear embedding residual block (LLERB) in the structural preservation phase can search for similar sparse patches from the PAN image space and embed the corresponding local geometric relationship into the residual space to enhance the MS image. Extensive experiments suggest that the proposed LLERN outperforms state-of-the-art methods from visual and quantitative perspectives, and confirm the assumption that LR image patches and residual image patches in a local region share a similar manifold structure, which can be used to guide deep-learning modeling with improved interpretability. The source code is available at https://github.com/jiaming-wang/LLERN. [ABSTRACT FROM AUTHOR]

Object-based convolutional neural networks (OCNNs) have achieved great performance in the field of land-cover and land-use classification. Studies have suggested that the generation of object convolutional positions (OCPs) largely determines the performance of OCNNs. Optimized distribution of OCPs facilitates the identification of segmented objects with irregular shapes. In this study, we propose a morphology-based binary tree sampling (BTS) method that provides a reasonable, effective, and robust strategy to generate evenly distributed OCPs. The proposed BTS algorithm consists of three major steps: 1) calculating the required number of OCPs for each object, 2) dividing a vector object into smaller sub-objects, and 3) generating OCPs based on the sub-objects. Taking the object identification in land-cover and land-use classification as a case study, we compare the proposed BTS algorithm with other competing methods. The results suggest that the BTS algorithm outperforms all other competing methods, as it yields more evenly distributed OCPs that contribute to better representation of objects, thus leading to higher object identification accuracy. Further experiments suggest that the efficiency of BTS can be improved when multi-thread technology is implemented. [ABSTRACT FROM AUTHOR]

Due to the bird's eye view of remote sensing sensors, the orientational information of an object is a key factor that has to be considered in object detection. To obtain rotating bounding boxes, existing studies either rely on rotated anchoring schemes or adding complex rotating ROI transfer layers, leading to increased computational demand and reduced detection speeds. In this study, we propose a novel internal-external optimized convolutional neural network for arbitrary orientated object detection in optical remote sensing images. For the internal optimization, we designed an anchor-based single-shot head detector that adopts the concept of coarse-to-fine detection for two-stage object detection networks. The refined rotating anchors are generated from the coarse detection head module and fed into the refining detection head module with a link of an embedded deformable convolutional layer. For the external optimization, we propose an IOU balanced loss that addresses the regression challenges related to arbitrary orientated bounding boxes. Experimental results on the DOTA and HRSC2016 benchmark datasets show that our proposed method outperforms selected methods. [ABSTRACT FROM AUTHOR]

Face hallucination is a domain-specific super-resolution (SR) algorithm, that generates high-resolution (HR) images from the observed low-resolution (LR) inputs. Recently, deep convolutional neural network (CNN) based SR offers an end-to-end solution for learning the complex relationship between LR and HR images, and achieves superior performance. However, most of them ignore the role of high-frequency (HF) information in image recovering. We design a novel global-local fused network (GLFSR) to refine HF information for recovering fine details of facial images. In contrast to existing methods that often increase the depth of network, we enhance the residual HF information from local to global levels through the networks. The proposed global-local fused network involves four sub-modules: first, reconstruction network, which is used to super-resolve the synthetic HR image from pixel level by reconstruction network, local and global residual enhancement networks, which generate residual information for learning; and fusion module, which is used to generate the final HR image. Experimental results on CAS-PEAL-R1 and CASIA-Webface databases demonstrate that GLFSR is superior to other state-of-the-art deep learning approaches. [ABSTRACT FROM AUTHOR]

Deep learning has obtained state-of-the-art results in a variety of computer vision tasks and has also been used to solve SAR image classification problems. Deep learning algorithms typically require a large amount of training data to achieve high accuracy. In contrast, the size of SAR image datasets is often comparatively limited. Therefore, this paper proposes a novel method, deep memory convolution neural networks (M-Net), to alleviate the problem of overfitting caused by insufficient SAR image samples. Based on the convolutional neural networks (CNN), M-Net adds an information recorder to remember and store samples’ spatial features, and then it uses spatial similarity information of the recorded features to predict unknown sample labels. M-Net's use of this information recorder may cause difficulties for convergence if conventional CNN training methods were directly used to train M-Net. To overcome this problem, we propose a transfer parameter technique to train M-Net in two steps. The first step is to train a CNN, which has the same structure as the part of CNN incorporated in M-Net, to obtain initial training parameters. The second step applies the initialized parameters to M-Net and then trains the entire M-Net. This two-step training approach helps us to overcome the nonconvergence issue, and also reduces training time. We evaluate M-Net using the public benchmark MSTAR dataset, and achieve higher accuracy than several other well-known SAR image classification algorithms. [ABSTRACT FROM AUTHOR]

The deep learning algorithm has made great breakthroughs in optical image processing. Some deep learning algorithms require a large number of labeled samples for training. For PolSAR data sets, due to the influence of speckle noise and other factors, high-quality labeled data are limited. Therefore, it is meaningful to use deep learning algorithm to solve PolSAR classification problem in limited labeled dataset. This paper proposes a spatial feature-based convolutional neural network (SF-CNN). The network adopts a dual-branch CNN structure. Both of the two branches have the same structure and share parameters. SF-CNN can receive more than one sample as input. SF-CNN's special structure can expand the original training set by combining different samples, and alleviate the problem of insufficient labeled training data in PolSAR image classification tasks. When training, SF-CNN maps high-dimensional PolSAR image to low-dimensional feature space. In low-dimensional feature space, SF-CNN enhances the ability of network to extract discriminative features by maximizing or minimizing the distance between feature centers of different classes. In order to dig up the relationship between the samples, the test sample features are compared with every training sample feature when testing. Finally, labels of test samples are determined by the comparison result. The result of SF-CNN in PolSAR image classification task is better than that of standard CNN. • This paper proposes a spatial feature-based convolutional neural network (SF-CNN). • SF-CNN adopts a dual-branch CNN structure. • Both of the two branches have the same structure and share parameters. • SF-CNN can receive more than one sample as input. • SF-CNN's special structure can expand the original training set. [ABSTRACT FROM AUTHOR]

Remote Sensing Super-Resolution (RS-SR) constitutes a pivotal component in the domain of remote sensing image analysis, aimed at enhancing the spatial resolution of low-resolution imagery. Recent advancements have seen deep learning techniques achieving substantial progress in the RS-SR field. Notably, Graph Neural Networks (GNNs) have emerged as a potent mechanism for processing remote sensing images, adept at elucidating the intricate inter-pixel relationships within images. Nevertheless, a prevalent limitation among existing GNN-based methodologies is their disregard for the high computational demands, which circumscribes their applicability in environments with limited computational resources. This paper introduces a streamlined RS-SR framework, leveraging a Multi-Scale Graph Attention Network (MSGAN), designed to effectively balance computational efficiency with high performance. The core of MSGAN is a novel multi-scale graph attention module, integrating graph attention block and multi-scale lattice block structures, engineered to comprehensively assimilate both localized and extensive spatial information in remote sensing images. This enhances the framework's overall efficacy and resilience in RS-SR tasks. Comparative experimental analyses demonstrate that MSGAN delivers competitive results against state-of-the-art methods while reducing parameter count and computational overhead, presenting a promising avenue for deployment in scenarios with limited computational resources. • We propose an efficient lightweight reconstruction network for remote sensing. • We develop a multi-scale graph attention module to extract multi-scale features. • We design a global feature fusion module to fusion global contextual information. • We achieve more efficient reconstruction performance with fewer network parameters. [ABSTRACT FROM AUTHOR]