Dual Network Research Articles

Biomimetic Dual Absorption-Adsorption Networked MXene Aerogel-Pump for Integrated Water Harvesting and Power Generation System.

Harvesting atmospheric water and converting it into electricity play vital roles in advancing next-generation energy conversion systems. However, the current water harvester systems suffer from a weak water capture ability and poor recyclability due to high diffusion barriers and low sorption kinetics, which significantly limit their practical application. Herein, we drew inspiration from the natural "Pump effect" observed in wood and successfully developed a dual "absorption-adsorption" networked MXene aerogel atmospheric water harvester (MAWH) through ice templating and confining LiCl processes, thereby serving multiple purposes of clean water production, passive dehumidification, and power generation. The MAWH benefits from the dual H-bond network of MXene and cellulose nanocrystals (absorption network) and the hygroscopic properties of lithium chloride (adsorption network). Furthermore, its aligned wood-like channel structure efficiently eliminates water nucleation near the 3D network, resulting in fast moisture absorption. The developed MAWH demonstrates a high moisture absorption ability of 3.12 g g-1 at 90% relative humidity (RH), featuring rapid vapor transport rates and durable cyclic performance. When compared with commercial desiccants such as the 4A molecular sieve and silica gel, the MAWH can reduce the RH from 80% to 20% within just 6 h. Most notably, our integrated MAWH-based water harvesting-power generation system achieves a high voltage of ∼0.12 V at 77% RH, showcasing its potential for practical application. These developed MAWHs are considered as high-performance atmospheric water harvesters in the water collection and power generation field.

Preparation and characterization of peach gum/chitosan polyelectrolyte composite films with dual cross-linking networks for antibacterial packaging

Peach gum (PG) is a valuable polymeric feedstock for developing eco-friendly, bio-safe, and functional materials. However, PG has limited use in food packaging due to its inferior mechanical and antibacterial properties. To overcome these limitations, we created a dual cross-linked network by introducing chitosan (CS) and glycerol to the PG matrix. Our research discovered that incorporating CS into the PG matrix significantly improved its Young's modulus, from 277.62 to 925.89 MPa, and its tensile strength from 5.96 to 39.94 MPa. Furthermore, the inclusion of glycerol greatly increased the elongation. These enhancements were attributed to the ionic and hydrogen-bonding interactions between the two biopolymers. Additionally, the composite films exhibited strong antibacterial effects, reducing the total number of colonies by 99.2 % and 99.9 % against Escherichia coli (E. coli) and Staphylococcus aureus (S. aureus), respectively. The incorporation of CS resulted in more amorphous films, enhancing their stiffness, flexibility, and barrier properties. To assess the practical application of PG/CS composite films, we conducted a comparative analysis between non-packaged strawberries and strawberries packaged with these films. The results demonstrated that the composite polyelectrolyte film extended the shelf life of strawberries better than the non-packaged fruits.

Context-detail-aware United Network for Single Image Deraining

Images captured outdoors are often affected by rainy days, resulting in a severe deterioration in the visual quality of the captured images and a decrease in the performance of related applications. Therefore, single image deraining has attracted attention as a challenging research topic. Nowadays, there are two common deraining architectures in single image deraining. The first one is to restore the rain-free image by deducting rain streaks learned by the model from the rain image, but the background structure is easily mistaken for rain streaks and subtracted. The other one is to directly learn the clean background structure through the model using rain images, but it is difficult to completely remove the rain streaks due to the complexity of the information in images with rain. Therefore, current methods cannot balance rain streak removal and rain-free image background restoration in a single architecture and achieve good results. To address this issue, we propose a novel framework, namely, Context-Detail-Aware United Network (CDaUNet), which combines the above two architectures in this study. More specifically, we divide the restoration of the background structure of rain-free images and the learning of rain streaks into two independent sub-networks. The proposed Structure-Aware Rain Removal Network (SaRRN) is to learn the background structure in images to reconstruct clean rain-free images, whereas Detail-Aware Rain Streak Learning Network (DaRLN) is proposed to learn the details of rain streaks in images. Finally, we fuse the results generated by the two sub-networks through our designed Dual Architecture Fusion Network (DAFN) to reconstruct original rain images to effectively fuse the results of the two sub-networks. The experimental results show that CDaUNet achieves satisfactory performance in comparison with the state-of-the-art approaches included in rain streak removal and rain-free image structure restoration architectures on both synthetic and real image datasets, confirming the effectiveness of our method.

Attention based dual UNET network for infrared and visible image fusion

Attention based dual UNET network for infrared and visible image fusion

AB2CD: AI for Building Climate Damage Classification and Detection

We explore the implementation of deep learning techniques for precise building damage assessment in the context of natural hazards, utilizing remote sensing data. The xBD dataset, comprising diverse disaster events from across the globe, serves as the primary focus, facilitating the evaluation of deep learning models. We tackle the challenges of generalization to novel disasters and regions while accounting for the influence of low-quality and noisy labels inherent in natural hazard data. Furthermore, our investigation quantitatively establishes that the minimum satellite imagery resolution essential for effective building damage detection is 3 meters and below 1 meter for classification using symmetric and asymmetric resolution perturbation analyses. To achieve robust and accurate evaluations of building damage detection and classification, we evaluated different deep learning models with residual, squeeze and excitation, and dual path network backbones, as well as ensemble techniques. Overall, the U-Net Siamese network ensemble with F-1 score of 0.812 performed the best against the xView2 challenge benchmark. Additionally, we evaluate a Universal model trained on all hazards against a flood expert model and investigate generalization gaps across events, and out of distribution from field data in the Ahr Valley. Our research findings showcase the potential and limitations of advanced AI solutions in enhancing the impact assessment of climate change-induced extreme weather events, such as floods and hurricanes. These insights have implications for disaster impact assessment in the face of escalating climate challenges.

Fabrication and Evaluation of an Alginate-Based Dual Network Mulch Film Reinforced with Multiscale Cellulose Fibers

Fabrication and Evaluation of an Alginate-Based Dual Network Mulch Film Reinforced with Multiscale Cellulose Fibers

Synthetic Document Images with Diverse Shadows for Deep Shadow Removal Networks.

Shadow removal for document images is an essential task for digitized document applications. Recent shadow removal models have been trained on pairs of shadow images and shadow-free images. However, obtaining a large, diverse dataset for document shadow removal takes time and effort. Thus, only small real datasets are available. Graphic renderers have been used to synthesize shadows to create relatively large datasets. However, the limited number of unique documents and the limited lighting environments adversely affect the network performance. This paper presents a large-scale, diverse dataset called the Synthetic Document with Diverse Shadows (SynDocDS) dataset. The SynDocDS comprises rendered images with diverse shadows augmented by a physics-based illumination model, which can be utilized to obtain a more robust and high-performance deep shadow removal network. In this paper, we further propose a Dual Shadow Fusion Network (DSFN). Unlike natural images, document images often have constant background colors requiring a high understanding of global color features for training a deep shadow removal network. The DSFN has a high global color comprehension and understanding of shadow regions and merges shadow attentions and features efficiently. We conduct experiments on three publicly available datasets, the OSR, Kligler's, and Jung's datasets, to validate our proposed method's effectiveness. In comparison to training on existing synthetic datasets, our model training on the SynDocDS dataset achieves an enhancement in the PSNR and SSIM, increasing them from 23.00 dB to 25.70 dB and 0.959 to 0.971 on average. In addition, the experiments demonstrated that our DSFN clearly outperformed other networks across multiple metrics, including the PSNR, the SSIM, and its impact on OCR performance.

Highly stretchable and room-temperature self-healing sheath-core structured composite fibers for ultrasensitive strain sensing and visual thermal management

Highly stretchable and self-healing wearable electronics for strain sensing and Joule heating are highly desirable for future emerging applications of wearable devices, smart robots, human-machine interface and artificial intelligence, etc. Herein, the highly stretchable and room-temperature self-healing sheath-core structured composite fibers are fabricated via the feasible in-situ polymerization modification followed by electroless silver-plating approach. Benefitting from the simultaneous incorporation of dual dynamic reversible chemical networks and construction of sheath-core structures, the composite fibers show ultrasensitive strain sensing and visual thermal management performances with excellent room-temperature self-healing capacity at ultralow Ag loadings. The composite fibers after cutting and self-healing also exhibit outstanding strain sensing performances with a high gauge factor (GF) of 64.0 and visual thermal management performances with tailorable Joule heating temperatures. Furthermore, they possess excellent working stability and reliability in practical applications of human motion detection and personal thermal management. This work demonstrates the fabrication of highly stretchable and room-temperature self-healing composite fibers for next-generation wearable devices, smart robots, human-machine interface and artificial intelligence, etc.

Secondary Chemical Cross-Linking to Improve Mechanical Properties in a Multifaceted Biocompatible Strain Sensor.

A new conductive and transparent organohydrogel is developed with high stretchability, excellent mechanical, self-healing, antifreezing, and adhesive properties. A simple one-pot polymerization method is used to create polyacrylamide cross-linked through N,N'-methylenebis(acrylamide) (MBAA) and divinylbenzene (DVB). The dual chemical cross-linked gel network is complemented by several physical cross-links via hydrogen bonding and π-π interaction. Multiple chemical and physical cross-links are used to construct the gel network that allows toughness (171 kPa), low modulus (≈45 kPa), excellent stretchability (>1100%), and self-healing ability. The use of appropriate proportions of the water/glycerol binary solvent system ensures efficient environment tolerance (-20 to 40 °C). Phytic acid is used as a conductive filler that provides excellent conductivity and contributes to the physical cross-linking. Dopamine is incorporated in the gel matrix, which endows excellent adhesive property of the gel. The organohydrogel-based strain sensors are developed with state-independent properties, highly linear dependence, and excellent antifatigue performance (>100 cycles). Moreover, during the practical wearable sensing tests, human motions can be detected, including speaking, smiling, and joint movement. Additionally, the sensor is biocompatible, indicating the potential applications for the next generation of epidermal sensors.

Dual generative adversarial networks based on regression and neighbor characteristics.

Imbalanced data is a problem in that the number of samples in different categories or target value ranges varies greatly. Data imbalance imposes excellent challenges to machine learning and pattern recognition. The performance of machine learning models leans to be partially towards the majority of samples in the imbalanced dataset, which will further affect the effect of the model. The imbalanced data problem includes an imbalanced categorical problem and an imbalanced regression problem. Many studies have been developed to address the issue of imbalanced classification data. Nevertheless, the imbalanced regression problem has not been well-researched. In order to solve the problem of unbalanced regression data, we define an RNGRU model that can simultaneously learn the regression characteristics and neighbor characteristics of regression samples. To obtain the most comprehensive sample information of regression samples, the model uses the idea of confrontation to determine the proportion between the regression characteristics and neighbor characteristics of the original samples. According to the regression characteristics of the regression samples, an index ccr (correlation change rate) is proposed to evaluate the similarity between the generated samples and the original samples. And on this basis, an RNGAN model is proposed to reduce the similarity between the generated samples and the original samples by using the idea of confrontation.

Dual residual attention network for image denoising

In image denoising, deep convolutional neural networks (CNNs) can obtain favorable performance on removing spatially invariant noise. However, many of these networks cannot perform well on removing the real noise (i.e. spatially variant noise) that is generated during image acquisition or transmission, which severely impedes their application in practical image denoising tasks. In this paper, we propose a novel Dual-branch Residual Attention Network (DRANet) for image denoising, which has both the merits of a wide model architecture and the attention-guided feature learning. The proposed DRANet includes two different parallel branches, which can capture complementary features to enhance the learning ability of the model. We designed a new residual attention block (RAB) and a novel hybrid dilated residual attention block (HDRAB) for the upper and lower branches, respectively. The RAB and HDRAB can capture rich local features through multiple skip connections between different convolutional layers, and the unimportant features can be dropped. Meanwhile, the long skip connections in each branch and the global feature fusion between the two parallel branches can effectively capture the global features as well. Extensive experiments demonstrate that compared with other state-of-the-art denoising methods, our DRANet can produce competitive denoising performance both on the synthetic and real-world noise removal. The code for DRANet is accessible at https://github.com/WenCongWu/DRANet.

An interpretable dual attention network for diabetic retinopathy grading: IDANet

Diabetic retinopathy (DR) is the most prevalent cause of visual impairment in adults worldwide. Typically, patients with DR do not show symptoms until later stages, by which time it may be too late to receive effective treatment. DR Grading is challenging because of the small size and variation in lesion patterns. The key to fine-grained DR grading is to discover more discriminating elements such as cotton wool, hard exudates, hemorrhages, microaneurysms etc. Although deep learning models like convolutional neural networks (CNN) seem ideal for the automated detection of abnormalities in advanced clinical imaging, small-size lesions are very hard to distinguish by using traditional networks. This work proposes a bi-directional spatial and channel-wise parallel attention based network to learn discriminative features for diabetic retinopathy grading. The proposed attention block plugged with a backbone network helps to extract features specific to fine-grained DR-grading. This scheme boosts classification performance along with the detection of small-sized lesion parts. Extensive experiments are performed on four widely used benchmark datasets for DR grading, and performance is evaluated on different quality metrics. Also, for model interpretability, activation maps are generated using the LIME method to visualize the predicted lesion parts. In comparison with state-of-the-art methods, the proposed IDANet exhibits better performance for DR grading and lesion detection.

Dual convolutional network based on hypergraph and multilevel feature fusion for road extraction from high-resolution remote sensing images

ABSTRACT Road extraction from high-resolution remote sensing images (HRSI) is confronted with the challenge that roads are occluded by other objects, including opaque obstructions and similarly colored areas. This paper proposes a dual convolutional network based on hypergraph and multilevel feature fusion (DHM) for road extraction to address these challenges. The DHM consists of two branch networks (HGNN branch and CNN branch) and a bimodal feature fusion module (BFFM). In the HGNN branch, an algorithm is developed to exploit the shape features of roads and construct hypergraphs on the HRSI. Then, hypergraph neural networks are employed for the first time to capture the long-range context of roads to enhance road connectivity. In the CNN branch, a bottleneck fusion module integrated with an encoder-decoder network structure is built to aggregate multiscale local features. In BFFM, the long-range context from the HGNN branch and the local features from the CNN branch are fused through the designed position converter and enhanced graph reasoning module to achieve the complementary advantages of the dual-branch network. Extensive experiments on three datasets show that DHM outperforms other state-of-the-art methods, especially on the GS-Mountain road dataset. Furthermore, DHM significantly improves road extraction in occluded and similar road regions.

Dual Network Hydrogels Based on PRP and SA Promote the Retention Rate and Vascularization of Transplanted Fat.

Autologous adipose tissue often experiences ischemia and hypoxia after transplantation, leading to low retention rates and unstable operative impacts due to necrotic absorption. Platelet-rich plasma (PRP) can enhance fat regeneration and increase the fat retention rate after transplantation. However, the quick release of growth factors (GFs) in PRP decreases therapeutic efficiency. This study aimed to achieve a slow release of PRP to promote fat retention. We prepared a dual-network hydrogel (DN gel) based on FDA-approved PRP and sodium alginate (SA) through a simple "one-step" activation process. In vivo study, adipose tissue with saline (control group), SA gel (SA gel group), PRP gel (PRP gel group), and DN gel (DN gel group) was injected subcutaneously into the dorsum of nude mice. At 4 and 12 weeks after injection, tissues were assessed for volume and weight. Hematoxylin and eosin staining (HE) and immunofluorescence staining were performed for histological assessment. DN gel exhibits long-lasting growth factor effects, surpassing conventional clinical PRP gel regarding vascularization potential. In fat transplantation experiments, DN gel demonstrated improved vascularization of transplanted fat and increased retention rates, showing promise for clinical applications. DN gel-assisted lipofilling can significantly improve the retention rate and quality of transplanted fat. DN gel-assisted lipofilling, which is considered convenient, is a promising technique to improve neovascularization and fat survival. This journal requires that authors assign a level of evidence to each submission to which Evidence-Based Medicine rankings are applicable. This excludes Review Articles, Book Reviews, and manuscripts that concern Basic Science, Animal Studies, Cadaver Studies, and Experimental Studies. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266.

Knowledge-based Dual External Attention Network for peptide detectability prediction

Accurate prediction of peptide detectability plays a crucial role in various proteomics data analyses such as peptide identification and protein quantification. To improve the precision of peptide detectability prediction, we propose a novel approach called the Knowledge-based Dual External Attention Network (KDEAN). KDEAN introduces several innovative elements to enhance its representation and prediction capabilities. Firstly, it extracts valuable knowledge-based features from peptide sequences to facilitate the pattern recognition process. Secondly, KDEAN adopted dual networks to separately train peptide sequences from both the forward and backward directions, capturing comprehensive information. Thirdly, an external attention mechanism is utilized to identify and understand the connections between different peptide samples. The structure of KDEAN enables long-term dependency learning from both directions of the peptide sequences. Extensive evaluations on four testing datasets demonstrate that KDEAN outperforms existing methods, achieving a higher average performance in peptide detectability prediction. Additionally, comprehensive ablation studies confirm the effectiveness and advantages of key components in KDEAN, including knowledge-based feature representation, dual network architecture, and external attention.

Few-shot learning based oral cancer diagnosis using a dual feature extractor prototypical network

A large global health issue is cancer, wherein early diagnosis and treatment have proven to be life-saving. This holds true for oral cancer, thus emphasizing the significance of timely intervention. Deep learning techniques have gained traction in early cancer detection, exhibiting promising outcomes in accurate diagnosis. However, collecting a substantial amount of training data poses a challenge for deep learning models in cancer diagnosis. To address this limitation, this study proposes an oral cancer diagnosis approach based on a few-shot learning framework that circumvents the need for extensive training data. Specifically, a prototypical network is employed to construct a diagnostic model, wherein two feature extractors are utilized to extract prototypical features and query features respectively, departing from the conventional use of a single feature extraction function in prototypical networks. Moreover, a customized loss function is designed for the proposed method. Rigorous experimentation using a histopathological image dataset demonstrates the superior performance of our proposed approach over comparison methods.

Valorization of waste plastics to prepare tough, stretchable and highly conductive ionicgels

Polyethylene terephthalate (PET) is one of the most widely used polymers, and its intensive use has resulted in significant and growing environmental problems. The urgent need for a more efficient approach that avoids the cumbersome depolymerisation process and monomer purification to produce high-value products is evident. In this study, a simple and one-pot strategy was used to obtain polyurethane pre-polymer through the glycolysis of PET. The polyurethane network and the second polymer network formed by the monomers polyacrylic acid and polyacrylamide were co-introduced into the ionicgels to create a dual network structure. Glycolysis of PET was catalyzed using the ionic liquid catalyst [EMIM][DCA], resulting in a 100 % conversion. The PET alcoholysis solution consisted of 90.47 % BHET, 9.53 % oligomers containing benzene rings, and unreacted glycol. Polyurethanes synthesized through the one-pot method using PET hydrolysate, with benzene ring-containing BHETs and oligomers as the hard domains and ethylene glycol as the soft domains, exhibited excellent mechanical properties. PET-based polyurethane was employed in the preparation of dual network-structured ionicgels. As a result, the ionic gels exhibited a good stress–strain of 4.2 MPa and 990 %, along with a high conductivity (0.09 S/cm) and ultrahigh sensitivity. Furthermore, it could easily repair the broken ionicgels at 25 ℃ for 6 h, thus restoring their original tensile, and sensing properties.

Classification of Parkinson’s disease EEG signals using 2D-MDAGTS model and multi-scale fuzzy entropy

Parkinson’s disease (PD) is a neurodegenerative disorder that causes changes in neurons, behavior, and physiological structures. However, these changes are very subtle in the early stages of PD, making diagnosis and treatment challenging. To overcome this challenge, we propose a multi-scale fuzzy entropy (MSFEn) fusion method. MSFEn can be used as an EEG feature to quantify the complexity and irregularity of the EEG signal. We also propose a two-dimensional multiple dual attention gated temporal-separable (2D-MDAGTS) model for PD automatic detection. This model integrates temporal separable convolution (TSCN), gated recurrent unit (GRU), and dual attention network (DANet) to improve PD detection performance. TSCN can mine multi-level information in timing sequence and output processed features. Then the GRU is operated in parallel with the DANet to further integrate feature information. GRU can capture long-term dependencies of time series data and DANet can adjust the weight of features through the attention mechanism to better focus on the features related to PD. Two datasets were used to evaluate the proposed methods. In the classification of healthy subjects and drug-free PD patients, our results achieved an accuracy of 98.68% on the San Diego dataset and 99.30% on the UNM dataset. In the classification of healthy subjects and PD patients on medication, our results achieved an accuracy of 99.01% on the San Diego dataset and 99.31% on the UNM dataset. The results showed that this method is superior in the diagnosis of PD patients. The application of this method is expected to provide support for early diagnosis and disease monitoring.

An early warning approach for the rapid identification of extreme weather disasters based on phased array dual polarization radar cooperative network data.

In recent years, X-band phase array dual polarization weather radar technology has matured. The cooperative networking data from X-band phase array dual polarization weather radar have many advantages compared with traditional methods, namely, high spatial and temporal resolution (approximately 70 seconds in one scan, 30 m in radial distance resolution), wide coverages that can compensate for the observation blind spots, and data fusion technology that is used in the observation overlap area to ensure that the observed precipitation data have spatial continuity. Based on the above radar systems, this study proposes an improved hail and lightning weather disaster rapid identification and early warning algorithm. The improved thunderstorm identification, tracking, analysis, and nowcasting (TITAN) algorithm is used to quickly identify three-dimensional strong convective storm cells. Large sample observation experiment data are used to invert the localized hail index (HDR) to identify the hail position. The fuzzy logic method is used to comprehensively determine the probability of lightning occurrence. The comparative analysis experiment shows that, compared with the live observation data from the ground-based automatic station, the hail and lightning disaster weather warning algorithm developed by this study can increase warning times by approximately 7 minutes over the traditional algorithm, and its critical success index (CSI), false alarm ratio (FAR) and omission alarm ratio (OAR) scores are better than those of the traditional method. The average root mean square error (ARMSE) for identifying hail and lightning locations by this improved method is also significantly better than that of traditional methods. We show that our method can provide probabilistic predictions that improve hail and lightning identification, improve the precision of early warning and support operational utility at higher resolutions and with greater lead times that traditional methods struggle to achieve.

Rapidly recyclable, monomer recovery and flame-retardant bio-based polyimine networks

With the development of a circular economy and sustainable environment, efficient and sustainable recycling of useful chemicals from thermosets is of great significance. However, most commercial thermosetting materials are difficult to reprocess, degrade, and recycle due to their permanent cross-linked network, and the preparation process consumes much fossil energy and causes serious environmental pollution problems. Furthermore, the high flammability makes them unsafe during service life and limits their applications. Herein, we used renewable lignin derivative vanillin as bio-based materials to synthesize a trialdehyde monomer (TMP) with phosphorus element and developed the bio-based polyimine dynamic networks with excellent comprehensive properties including outstanding reprocess ability, degradation recyclability, and fire resistance performance based on polyetherimide and 4-aminophenyl disulfide. The prepared bio-based polyimine materials with the dual dynamic networks (BPDNs) exhibited outstanding mechanical property with tensile strength of 61.6 MPa, super-rapid degradation rate and high monomer recovery rates of 70.5 %. The prepared BPDNs samples showed great flame retardancy with V0 rating and V1 rating in UL-94 vertical burning test and high LOI value of 29 %. Moreover, BPDNs could be reprocessed within 10 min at elevated temperatures (150 °C). This work provides a very effective method to prepare bio-based advanced thermosetting materials with multiple functions.