Dual Network Research Articles

Achieving Better Category Separability for Hyperspectral Image Classification: A Spatial-Spectral Approach.

The task of hyperspectral image (HSI) classification has attracted extensive attention. The rich spectral information in HSIs not only provides more detailed information but also brings a lot of redundant information. Redundant information makes spectral curves of different categories have similar trends, which leads to poor category separability. In this article, we achieve better category separability from the perspective of increasing the difference between categories and reducing the variation within category, thus improving the classification accuracy. Specifically, we propose the template spectrum-based processing module from spectral perspective, which can effectively expose the unique characteristics of different categories and reduce the difficulty of model mining key features. Second, we design an adaptive dual attention network from spatial perspective, where the target pixel can adaptively aggregate high-level features by evaluating the confidence of effective information in different receptive fields. Compared with the single adjacency scheme, the adaptive dual attention mechanism makes the ability of target pixel to combine spatial information to reduce variation more stable. Finally, we designed a dispersion loss from the classifier's perspective. By supervising the learnable parameters of the final classification layer, the loss makes the category standard eigenvectors learned by the model more dispersed, which improves the category separability and reduces the rate of misclassification. Experiments on three common datasets show that our proposed method is superior to the comparison method.

A dual domain multi-exposure image fusion network based on spatial-frequency integration

Multi-exposure image fusion aims to generate a single high-dynamic image by integrating images with different exposures. Existing deep learning-based multi-exposure image fusion methods primarily focus on spatial domain fusion, neglecting the global modeling ability of the frequency domain. To effectively leverage the global illumination modeling ability of the frequency domain, we propose a novelty perspective on multi-exposure image fusion via the Spatial-Frequency Integration Framework, named MEF-SFI. Initially, we revisit the properties of the Fourier transform on the 2D image, and verify the feasibility of multi-exposure image fusion in the frequency domain where the amplitude and phase component is able to guide the integration of the illumination information. Subsequently, we present the deep Fourier-based multi-exposure image fusion framework, which consists of a spatial path and frequency path for local and global modeling separately. Specifically, we introduce a Spatial-Frequency Fusion Module to facilitate efficient interaction between dual domains and capture complementary information from input images with different exposures. Finally, we combine a dual domain loss function to ensure the retention of complementary information in both the spatial and frequency domains. Extensive experiments on the PQA-MEF dataset demonstrate that our method achieves visual-appealing fusion results against state-of-the-art multi-exposure image fusion approaches. Our code is available at https://github.com/SSyangguang/MEF-freq.

Fast Convergent Antinoise Dual Neural Network Controller With Adaptive Gain for Flexible Endoscope Robots.

Manual rigid endoscopes have defects such as a low efficiency, difficult operation, and safety risks, and the antinoise interference ability, convergence speed, and control accuracy of the neural network control technology for the existing autonomous endoscopes are often ignored. Solving these problems is important for the stable operation of endoscopes. Therefore, a new adaptive fast convergent antinoise dual neural network (AFA-DNN) controller for the visual servo control of ten-degree of freedom flexible endoscope robots (FERs) with physical constraints is proposed in this work. First, the control scheme of the FERs is formulated as a quadratic programming problem, and then, an AFA-DNN visual servo controller is designed for the FERs. The adaptive gains of the controller can accelerate the convergence, improve the antinoise ability, and increase the convergence accuracy of the controller. Then, according to the Lyapunov theory, the fast convergence of the AFA-DNN in finite time is proven for both noise-free and noisy conditions. The experimental results indicate that the FER controlled by the proposed AFA-DNN can accurately track various trajectories and that the AFA-DNN has a better antinoise interference ability, higher convergence accuracy, and faster convergence speed than conventional methods. The convergence speed of the AFA-DNN is increased by a factor of 4.22 by using the adaptive gains. Experiments also indicate that the AFA-DNN remains well functioning under various noise disturbances (such as constant, periodic, linear, and Gaussian noise).

DualNetGO: a dual network model for protein function prediction via effective feature selection.

Protein-protein interaction (PPI) networks are crucial for automatically annotating protein functions. As multiple PPI networks exist for the same set of proteins that capture properties from different aspects, it is a challenging task to effectively utilize these heterogeneous networks. Recently, several deep learning models have combined PPI networks from all evidence, or concatenated all graph embeddings for protein function prediction. However, the lack of a judicious selection procedure prevents the effective harness of information from different PPI networks, as these networks vary in densities, structures, and noise levels. Consequently, combining protein features indiscriminately could increase the noise level, leading to decreased model performance. We develop DualNetGO, a dual-network model comprised of a Classifier and a Selector, to predict protein functions by effectively selecting features from different sources including graph embeddings of PPI networks, protein domain, and subcellular location information. Evaluation of DualNetGO on human and mouse datasets in comparison with other network-based models shows at least 4.5%, 6.2%, and 14.2% improvement on Fmax in BP, MF, and CC gene ontology categories, respectively, for human, and 3.3%, 10.6%, and 7.7% improvement on Fmax for mouse. We demonstrate the generalization capability of our model by training and testing on the CAFA3 data, and show its versatility by incorporating Esm2 embeddings. We further show that our model is insensitive to the choice of graph embedding method and is time- and memory-saving. These results demonstrate that combining a subset of features including PPI networks and protein attributes selected by our model is more effective in utilizing PPI network information than only using one kind of or concatenating graph embeddings from all kinds of PPI networks. The source code of DualNetGO and some of the experiment data are available at: https://github.com/georgedashen/DualNetGO.

A thiolated oxidized guar gum and sodium alginate dual-network microspheres with enhanced gastric acid resistance and mucoadhesion for delivery of probiotics

Probiotics offer numerous beneficial functions for human bodies, while the low survival rate under gastric acid and short retention time in the intestine are the major obstacles to their utilization. To address these issues, we designed a novel dual-network hydrogel microsphere that combines gastric acid resistance with enhanced mucoadhesion, aiming for the targeted delivery of probiotics. Thiolated oxidized guar gum (SOGG) was disulfide-linked to form the first network, and sodium alginate (SA) was cross-linked with Ca2+ to form the second network. Under the protection of the interpenetrating dual network microspheres, a much higher viability of Lactobacillus rhamnosus (LGG) (8.73 log CFU/mL) was achieved in simulated gastric fluid, compared to the zero-survival rate of free LGG. Mucoadhesion tests showed that the adhesion rate of SOGG/SA microspheres to the intestinal mucosa was 1.75 times higher than that of thiol-free microspheres. In vivo studies revealed that LGG-loaded microspheres significantly enhanced intestinal barrier function, remodeled the gut microbiome, and alleviated DSS-induced colitis in mice. Overall, SOGG/SA microspheres provide an effective strategy to the challenges of probiotic reduction in the stomach and rapid expulsion from the intestines, enhancing their health benefits.

Fabrication of mechanically stable fabric-based hydrogel membranes based on dual network toughening strategy for high efficient oil–water emulsion separation

Massive discharge of oily wastewater originating from oil-related industry, residential sewage and frequent oil spill accidents leads to serious environmental problems and health concerns. Amphoteric polyelectrolytes and hydrogel materials have been employed in the context of oil–water separation to resist oil pollution. However, the weak mechanical properties limit their practical application in complex environments. To address this, the design of a mechanically stable superhydrophilic/underwater superoleophobic composite membrane (PSBMA/PAM@CF) was been proposed by ingeniously in-situ growth of 3- [N, N-dimethyl −[2-(2-methylpropyl-2-enoxyloxy) ethyl] ammonium] propane-1-sulfonate (PSBMA)/polyacrylamide (PAM) dual network (DN) hydrogel on cotton fabric. The DN hydrogel strategy increased the stress of PSBMA hydrogel from 0.200 Mpa to 5.80 Mpa. The toughening effect of the DN hydrogel and its entanglement with cotton fibers endowed the membranes with excellent mechanical properties, and resulted in stable underwater superoleophobic and antifouling properties even after mechanical stability test. The PSBMA/PAM@CF membrane exhibited a separation efficiency exceeding 98.0 % for all types of oil-in-water emulsions stabilized by surfactants, and the membrane’s permeance remained at 1.75 × 103 L·m−2·h−1·bar−1 after 30 cycles. Additionally, the membrane exhibited a high permeance recovery rate of 98.2 %, surpassing general membranes. More significantly, we visualized the separation process through fluid dynamics simulations and revealed the separation mechanism of oil–water emulsions by membranes. It is expected that this work will provide new prospects for designing anti-oil-fouling membranes and applying in practical oil/water separation.

CATFSID: A few-shot human identification system based on cross-domain adversarial training

With the advancement of wireless sensing technology, human identification based on WiFi sensing has garnered significant attention in the fields of human–computer interaction and home security. Despite the initial success of WiFi sensing based human identification when the environment is fixed, the performance of the trained identity sensing model will be severely degraded when applied to unfamiliar environments. In this paper, a cross-domain human identification system (CATFSID) is proposed, which is able to achieve environment migration of trained model using up to 3-shot. CATFSID utilizes a dual adversarial training network, including cross-adversarial training between source and source domain classifiers, and adversarial training between source and target domain discriminators to extract environment-independent identity features. Introducing a method based on pseudo-label prediction, which assigns labels to target domain samples similar to the source domain samples, reduces the distribution bias of identity features between the source and target domains. The experimental results show accuracy of 90.1% and F1-Score of 89.33% when using 3 samples per user in the new environment.

Stacked deformable convolution network with weighted non-local attention and branch residual connection for image quality assessment

Convolutional neural networks are data-driven Image Quality Assessment (IQA) models based on the convolutional operation in a rectangular window. Deformable convolutions with learnable receptive fields can efficiently extract structural features of irregular objects in an image. By superimposing deformable convolutions, a plug-and-play module is designed to obtain information for irregular geometric shapes. To selectively fuse shallow and deep features, we propose a weighted non-local attention (WNLA) module with the input and output of self-attention in a weighted manner. This paper proposes a dual branch residual full-reference IQA network that combines weighted non-local attention and stacked deformable convolution. The proposed network was trained on PIPAL dataset and tested on LIVE and TID2013. The cross-dataset evaluation shows that the network has a competitive generalization ability. Ablation experiments indicate that the proposed modules can effectively improve the performance of the network. Comparative experiments reveal that our network is superior to existing excellent networks. The codes for training, test and visualization are available at: https://github.com/Pengchang-haha/SDCN.git.

Unsupervised Anomaly Detection of Industrial Images Based on Dual Generator Reconstruction Networks

At present, deep learning techniques are increasingly utilized in computer vision and anomaly detection. To address the limitations of inadequate reconstruction capability and subpar performance in reconstruction-based anomaly detection, this study enhances the existing algorithm and introduces an unsupervised anomaly detection of industrial images algorithm based on dual generator reconstruction networks-DGRNet. The network consists of two generators and a discriminator, introducing a widely recognized denoising diffusion probabilistic model (DDPM) as one of the generators, an autoencoder (AE) as the other generator, and a decoder as the discriminator. The model is tested on the MVTec AD dataset, and in the case of no additional training data,the anomaly detection AUC result of DGRNet exceeds the baseline method based on reconstruction by 19.6 percentage points. The experimental results show that DGRNet can improve the detection performance in the anomaly detection algorithm based on unsupervised and reconstructed networks.

A Cellulose Reinforced Polyacrylamide/Polyvinyl Alcohol Dual Network Hydrogel for Flexible Sensors

AbstractIn this paper, polyacrylamide (PAM) was used as the main matrix material to form the first layer network structure, and polyvinyl alcohol (PVA) was used as the secondary matrix material to form the second layer network structure. The hydrogels were prepared by free radical in‐situ polymerization and physical cross‐linking. This hydrogel was characterized not only by its semi‐interpenetrating network structure, in which cellulose nanocrystals (CNC) penetrate through the polyacrylamide network but also by the physically and chemically cross‐linked hybrid PAM/PVA double network. The structure and appearance of the hydrogels were characterized by scanning electron microscope and Fourier transform infrared spectrometer, and a series of tests were carried out to detect their mechanical properties, electrical properties and sensing properties. The results show that the hydrogel has good mechanical properties (the elongation at break can reach 600 % and the stress at break can reach 0.16 MPa), and the hydrogel also has good sensing and electrical properties. The addition of organic solvents makes the hydrogels have excellent antifreeze and water retention properties and also have excellent swelling properties. Therefore, the PAM/PVA/CNC dual network interpenetrating composite hydrogel has potential application value in the field of flexible sensors.

DualAD: Dual adversarial network for image anomaly detection⋆

AbstractAnomaly Detection, also known as outlier detection, is critical in domains such as network security, intrusion detection, and fraud detection. One popular approach to anomaly detection is using autoencoders, which are trained to reconstruct input by minimising reconstruction error with the neural network. However, these methods usually suffer from the trade‐off between normal reconstruction fidelity and abnormal reconstruction distinguishability, which damages the performance. The authors find that the above trade‐off can be better mitigated by imposing constraints on the latent space of images. To this end, the authors propose a new Dual Adversarial Network (DualAD) that consists of a Feature Constraint (FC) module and a reconstruction module. The method incorporates the FC module during the reconstruction training process to impose constraints on the latent space of images, thereby yielding feature representations more conducive to anomaly detection. Additionally, the authors employ dual adversarial learning to model the distribution of normal data. On the one hand, adversarial learning was implemented during the reconstruction process to obtain higher‐quality reconstruction samples, thereby preventing the effects of blurred image reconstructions on model performance. On the other hand, the authors utilise adversarial training of the FC module and the reconstruction module to achieve superior feature representation, making anomalies more distinguishable at the feature level. During the inference phase, the authors perform anomaly detection simultaneously in the pixel and latent spaces to identify abnormal patterns more comprehensively. Experiments on three data sets CIFAR10, MNIST, and FashionMNIST demonstrate the validity of the authors’ work. Results show that constraints on the latent space and adversarial learning can improve detection performance.

Noise2Variance: Dual networks with variance constraint for self‐supervised real‐world image denoising

AbstractImage denoising aims to restore a clean image from a noisy image. Traditional methods utilizing convolutional neural networks (CNN) for denoising are trained using pairs of noisy and clean images to comprehend the transformation from a noisy image to a clean one. However, the acquisition of such image pairs in real‐world scenarios presents a challenge. Hence, numerous self‐supervised denoising techniques have been developed that do not require clean images for training. This study demonstrates that a straightforward loss design, concentrating on variance, can effectively train a standard CNN denoiser in a self‐supervised fashion. A novel theoretical framework is introduced for training a basic CNN denoising model using three constraints: mean, variance, and augmentation. The variance constraint is crucial as it prevents the trained model from converging to trivial solutions such as identity or zero mapping. This theory provides valuable insights for the development of new self‐supervised denoising methods. Furthermore, a method that applies this theory to proposed dual networks is developed, which consist of two standard CNN models predicting both the clean image and the noise. This approach enhances model capacity during training while minimizing computational costs during inference. This method exemplifies the implementation of the variance constraint and introduces a data constraint for dual networks. Notably, the proposed method only assumes the presence of additive white noise, irrespective of the noise distribution. This minimal assumption enhances the model's robustness against noise with complex or unknown distributions in real‐world distorted images. Experimental results indicate that the proposed Noise2Variance method exhibits commendable performance on peak signal noise ratio and structural similarity metrics compared to existing self‐supervised denoising techniques. Visual comparison of results further substantiates the efficacy of the proposed method. A comparison of model complexity reveals that the method is efficient among the compared CNN‐based techniques.

Unsupervised domain adaptation multi-level adversarial learning-based crossing-domain retinal vessel segmentation

BackgroundThe retinal vasculature, a crucial component of the human body, mirrors various illnesses such as cardiovascular disease, glaucoma, and retinopathy. Accurate segmentation of retinal vessels in funduscopic images is essential for diagnosing and understanding these conditions. However, existing segmentation models often struggle with images from different sources, making accurate segmentation in crossing-source fundus images challenging. MethodsTo address the crossing-source segmentation issues, this paper proposes a novel Multi-level Adversarial Learning and Pseudo-label Denoising-based Self-training Framework (MLAL&PDSF). Expanding on our previously proposed Multiscale Context Gating with Breakpoint and Spatial Dual Attention Network (MCG&BSA-Net), MLAL&PDSF introduces a multi-level adversarial network that operates at both the feature and image layers to align distributions between the target and source domains. Additionally, it employs a distance comparison technique to refine pseudo-labels generated during the self-training process. By comparing the distance between the pseudo-labels and the network predictions, the framework identifies and corrects inaccuracies, thus enhancing the accuracy of the fine vessel segmentation. ResultsWe have conducted extensive validation and comparative experiments on the CHASEDB1, STARE, and HRF datasets to evaluate the efficacy of the MLAL&PDSF. The evaluation metrics included the area under the operating characteristic curve (AUC), sensitivity (SE), specificity (SP), accuracy (ACC), and balanced F-score (F1). The performance results from unsupervised domain adaptive segmentation are remarkable: for DRIVE to CHASEDB1, results are AUC: 0.9806, SE: 0.7400, SP: 0.9737, ACC: 0.9874, and F1: 0.8851; for DRIVE to STARE, results are AUC: 0.9827, SE: 0.7944, SP: 0.9651, ACC: 0.9826, and F1: 0.8326. ConclusionThese results demonstrate the effectiveness and robustness of MLAL&PDSF in achieving accurate segmentation results from crossing-domain retinal vessel datasets. The framework lays a solid foundation for further advancements in cross-domain segmentation and enhances the diagnosis and understanding of related diseases.

Self‐healing and cell adhesion properties of dynamic and photo‐crosslinking PEG‐based dual networks soft hydrogels

AbstractSelf‐healing involves an autonomic response that repairs damage to living systems. In this study, polyethylene glycol (PEG) underwent aldehyde‐functionalization using the Albrigh Goldman oxidation technique. Flexible, dynamic crosslink points to encourage self‐healing are generated by the formation of acylhydrazone bonds formed between long‐chain PEG dialdehyde and adipic acid dihydrazide. Furthermore, methacrylated PEG was synthesized to support the polymer matrix through dynamic covalent bonds. The hydrogels exhibited self‐healing within the first 4 hours at room temperature and after 24 h. hydrogels with mechanical performance close to the original were obtained. They have good tensile and elongation strength and are well‐suited for interaction with living cells.

The Role of Cultural Landscape in Shaping the Urban Spatial Context of Changzhou (1843–2021) from a Spatial Syntax Perspective

Changzhou is a city situated in the eastern region of China and serves as a pivotal juncture of the UNESCO World Cultural Heritage site, the Grand Canal of China. The study employed spatial syntax foreground and background network analysis to examine the cultural landscape characteristics and determinants of the urban spatial context in Changzhou, influenced by the Grand Canal, with a focus on rivers, streets, and social networks. The findings revealed that: (1) Foreground network construction: serving as the fundamental framework of urban space, street and river networks exhibited distinct functionalities in terms of structural configuration, functional organization, and potential impact. (2) Background network construction: social networks identified through spatial structure recognition demonstrated sustainable characteristics, as their layout was not directly correlated to the foreground network but contributed to the identification of the identity and the traits of social networks. (3) Dual network adaptability to contemporary urban space: it was imperative to redefine the role of river networks to augment their potential impact in shaping scenic leisure systems. This study provides novel insights into preserving and acknowledging the urban spatial context while offering valuable references for safeguarding regional cultural landscapes.

Dual separated attention-based graph neural network

Graph neural networks (GNNs) offer a viable solution to model the inter-dependencies among labeled and unlabeled samples in a semi-supervised manner. However, their performances can degrade dramatically when the number of labels is extremely limited, which is due to the limitations of typical graph convolutional network design in most existing GNNs, including over-smoothing, difficulty in extending the propagation step, and failure to preserve the distinctiveness of each node. To address the issues, we propose a dual separated attention-based graph neural network (DSA-GNN) to deal with label scarcity in semi-supervised node classification. Firstly, DSA-GNN decouples feature propagation from representation transformation to alleviate the problems of over-smoothing and overfitting. Secondly, DSA-GNN separates self-representation learning from neighbor-representation learning by two feature extractors with different learnable parameters. As a result, the commonality between connected nodes and the distinctiveness of each node can be both preserved. Thirdly, DSA-GNN incorporates an attention-based label propagation mechanism to refine the label prediction of each node, by aggregating label prediction among the neighborhood based on adaptive edge coefficients. The extensive experimental results on the real-world datasets demonstrate the superiority of DSA-GNN for semi-supervised node classification, especially when the observed labels are extremely limited.

DAFNet: A Dual Attention Fusion-Based Face Super-Resolution Network

With the development of convolutional neural networks (CNNs), deep learning-based face super-resolution (FSR) approaches have achieved remarkable results in recent years. However, existing FSR methods often rely on face prior knowledge, which greatly increases network complexity and computation. In this paper, we propose DAFNet, a dual attention fusion-based FSR network comprising numerous face attention fusion modules (FAFMs). FAFM is a residual structure containing attention mechanism, which is divided into feature branches and attention branches. The feature branch introduces an hourglass block to extract multi-scale information, while the attention branch incorporates channel attention and spatial attention in series. This design ensures that the network prioritizes important information and effectively recovers detailed facial features. Luminance-chrominance error loss and gradient loss are introduced to guide the training process. Additionally, adversarial loss and perceptual loss are incorporated to enhance the recovery of visually realistic face images. Notably, our method can produce clear faces at a high-scale factor of eight times without relying on any facial prior information, effectively reducing network complexity. Quantitative and qualitative experiments conducted on the CelebA and Helen datasets underscore the effectiveness of the proposed model.

D2BGAN: Dual Discriminator Bayesian Generative Adversarial Network for Deformable MR-Ultrasound Registration Applied to Brain Shift Compensation.

During neurosurgical procedures, the neuro-navigation system's accuracy is affected by the brain shift phenomenon. One popular strategy is to compensate for brain shift using intraoperative ultrasound (iUS) registration with pre-operative magnetic resonance (MR) scans. This requires a satisfactory multimodal image registration method, which is challenging due to the low image quality of ultrasound and the unpredictable nature of brain deformation during surgery. In this paper, we propose an automatic unsupervised end-to-end MR-iUS registration approach named the Dual Discriminator Bayesian Generative Adversarial Network (D2BGAN). The proposed network consists of two discriminators and a generator optimized by a Bayesian loss function to improve the functionality of the generator, and we add a mutual information loss function to the discriminator for similarity measurements. Extensive validation was performed on the RESECT and BITE datasets, where the mean target registration error (mTRE) of MR-iUS registration using D2BGAN was determined to be 0.75 ± 0.3 mm. The D2BGAN illustrated a clear advantage by achieving an 85% improvement in the mTRE over the initial error. Moreover, the results confirmed that the proposed Bayesian loss function, rather than the typical loss function, improved the accuracy of MR-iUS registration by 23%. The improvement in registration accuracy was further enhanced by the preservation of the intensity and anatomical information of the input images.

Reliability Analysis of the Freeze–Thaw Cycle of Aeolian Sand Concrete Based on a Dual Neural Network in Series Structure Failure Mode

Reliability Analysis of the Freeze–Thaw Cycle of Aeolian Sand Concrete Based on a Dual Neural Network in Series Structure Failure Mode

An adaptive dual graph convolution fusion network for aspect-based sentiment analysis

An adaptive dual graph convolution fusion network for aspect-based sentiment analysis