Robust Network Research Articles

Self-toughening and self-reinforcing of polypropylene via melt-volume-pulsation-induced crystalline network

To achieve low-cost self-toughening and self-reinforcing is highly desired for the application and recycling of polypropylene (PP) but remains unresolved. This work reports the construction of a robust crystalline network via a volume pulsation injection molding (VPIM) process to meet the challenge. In particular, it is found that melt volume pulsation under a normal packing pressure can induce the formation of high-content γ-lamellae in PP. These γ-lamellae can effectively connect α-lamellae to construct the robust crystalline network, which imparts PP with greatly enhanced mechanical properties. Moreover, the robustness of the crystalline network and the mechanical properties of PP can be easily regulated by tuning the pulsatile frequency. Compared with those of conventional injection-molded samples, impact strength, yield strength, and tensile modulus of VPIM samples can be enhanced by 584.70 %, 13.96 %, and 29.84 %, respectively. To reveal the processing-structure-properties relationships, the microstructure and mechanical properties of PP under different processing conditions were discussed in detail. Furthermore, a theoretical analysis of the mechanical energy of polymer melt during volume pulsation process was conducted. The results disclose that the frequency-dependent microstructure and properties of PP are closely related to the frequency-dependent mechanical energy gained by melt. Hence, this paper paves a new avenue toward the low-cost self-toughening and self-reinforcing of PP for value-added application.

Differential impact of Paenibacillus infection on the microbiota of Varroa destructor and Apis mellifera

The Western honey bee (Apis mellifera) is a vital agricultural pollinator whose populations are threatened by the parasitic mite Varroa destructor and associated pathogens. While the impact of Paenibacillus species on honey bees, particularly Paenibacillus larvae causing American foulbrood, is documented, their effect on the microbiota of Varroa mites remains unclear. This study aimed to investigate the influence of Paenibacillus sp. on the bacterial communities of Varroa mites and adult honey bees. We hypothesized that Paenibacillus sp. would significantly alter the microbiota of Varroa mites but have minimal effect on that of adult honey bees. Utilizing 16S rRNA sequencing data from a previous study, we reanalyzed samples categorized into four groups based on Paenibacillus sp. infection load: highly infected and lowly infected honey bees (A. mellifera) and mites (V. destructor). Infection status was determined by Paenibacillus sp. read counts, with more than three reads indicating high infection. Microbial diversity was assessed using alpha and beta diversity metrics. Co-occurrence networks were constructed to visualize bacterial community assemblies, and network robustness was evaluated through node addition and removal tests. Keystone taxa were identified based on eigenvector centrality and relative abundance. Highly infected Varroa mites exhibited a significant reduction in alpha diversity and a markedly different bacterial community composition compared to lowly infected mites (p < 0.05). Their bacterial co-occurrence networks showed decreased connectivity and robustness, indicating a disruptive effect of Paenibacillus sp. In contrast, adult honey bees displayed no significant differences in alpha diversity or network structure between highly and lowly infected groups (p > 0.05), suggesting a resilient microbiota. Keystone taxa analysis revealed fewer central species in highly infected Varroa mites, potentially impacting network stability. High Paenibacillus sp. infection is associated with significant alterations in the microbiota of Varroa mites, disrupting bacterial communities and potentially affecting mite physiology. The microbiota of adult honey bees appears more robust against Paenibacillus sp. influence. These findings enhance our understanding of the complex interactions within the "honey bee–mite–microorganism" system and may inform future strategies for managing Varroa mite infestations and associated pathogens.

Resiliency-informed optimal scheduling of smart distribution network with urban distributed photovoltaic: A stochastic P-robust optimization

An increased worldwide emphasis on resilience has resulted in a greater concentration on the distribution networks' resiliency. Therefore, the conventional approach to advance scheduling for power distribution networks is no longer sufficient, thereby requiring the creation of resilient scheduling approaches. This paper examines a proactive day-ahead scheduling method for distribution networks that are linked with significant renewable energy sources. The suggested approach employs a probabilistic scenario-generating strategy to quantify the uncertainties related to wind speed, solar irradiation, catastrophe duration, and load fluctuations. This study utilizes a hybrid stochastic p-robust optimization technique to enhance the analysis and provide robust strategies against risks. The present methodology employs a regret-based metric to quantify the adverse impacts of uncertainty inside the proactive scheduling procedure. The suggested methodology is verified using an IEEE 33-bus test system that includes significant distributed photovoltaic sources. The findings indicate a significant decrease of 43% in the highest level of regret, accompanied by a relatively limited rise of 1.22% in the operational expenses of the distribution network. This framework efficiently manages natural occurrences in the day-ahead scheduling process, making it a very promising method for improving the robustness of power distribution networks.

DeepArUco++: Improved detection of square fiducial markers in challenging lighting conditions

Fiducial markers are a computer vision tool used for object pose estimation and detection. These markers are highly useful in fields such as industry, medicine and logistics. However, optimal lighting conditions are not always available, and other factors such as blur or sensor noise can affect image quality. Classical computer vision techniques that precisely locate and decode fiducial markers often fail under difficult illumination conditions (e.g. extreme variations of lighting within the same frame). Hence, we propose DeepArUco++, a deep learning-based framework that leverages the robustness of Convolutional Neural Networks to perform marker detection and decoding in challenging lighting conditions. The framework is based on a pipeline using different Neural Network models at each step, namely marker detection, corner refinement and marker decoding. Additionally, we propose a simple method for generating synthetic data for training the different models that compose the proposed pipeline, and we present a second, real-life dataset of ArUco markers in challenging lighting conditions used to evaluate our system. The developed method outperforms other state-of-the-art methods in such tasks and remains competitive even when testing on the datasets used to develop those methods. Code available in GitHub: https://github.com/AVAuco/deeparuco/.

Understanding adversarial robustness against on-manifold adversarial examples

Deep neural networks (DNNs) are shown to be vulnerable to adversarial examples. A well-trained model can be easily attacked by adding small perturbations to the original data. One of the hypotheses of the existence of the adversarial examples is the off-manifold assumption: adversarial examples lie off the data manifold. However, recent researches showed that on-manifold adversarial examples also exist. In this paper, we revisit the off-manifold assumption and study a question: at what level is the poor adversarial robustness of neural networks due to on-manifold adversarial examples? Since the true data manifold is unknown in practice, we consider two approximated on-manifold adversarial examples on both real and synthesis datasets. On real datasets, we show that on-manifold adversarial examples have greater attack rates than off-manifold adversarial examples on both standard-trained and adversarially-trained models. On synthetic datasets, theoretically, we prove that on-manifold adversarial examples are powerful, yet adversarial training focuses on off-manifold directions and ignores the on-manifold adversarial examples. Furthermore, we provide analysis to show that the properties derived theoretically can also be observed in practice. Our analysis suggests that on-manifold adversarial examples are important. We should pay more attention to on-manifold adversarial examples to train robust models.

YOLO-Faster: An efficient remote sensing object detection method based on AMFFN.

As a pivotal task within computer vision, object detection finds application across a diverse spectrum of industrial scenarios. The advent of deep learning technologies has significantly elevated the accuracy of object detectors designed for general-purpose applications. Nevertheless, in contrast to conventional terrestrial environments, remote sensing object detection scenarios pose formidable challenges, including intricate and diverse backgrounds, fluctuating object scales, and pronounced interference from background noise, rendering remote sensing object detection an enduringly demanding task. In addition, despite the superior detection performance of deep learning-based object detection networks compared to traditional counterparts, their substantial parameter and computational demands curtail their feasibility for deployment on mobile devices equipped with low-power processors. In response to the aforementioned challenges, this paper introduces an enhanced lightweight remote sensing object detection network, denoted as YOLO-Faster, built upon the foundation of YOLOv5. Firstly, the lightweight design and inference speed of the object detection network is augmented by incorporating the lightweight network as the foundational network within YOLOv5, satisfying the demand for real-time detection on mobile devices. Moreover, to tackle the issue of detecting objects of different scales in large and complex backgrounds, an adaptive multiscale feature fusion network is introduced, which dynamically adjusts the large receptive field to capture dependencies among objects of different scales, enabling better modeling of object detection scenarios in remote sensing scenes. At last, the robustness of the object detection network under background noise is enhanced through incorporating a decoupled detection head that separates the classification and regression processes of the detection network. The results obtained from the public remote sensing object detection dataset DOTA show that the proposed method has a mean average precision of 71.4% and a detection speed of 38 frames per second.

A Fast and Robust Attention-Free Heterogeneous Graph Convolutional Network

A Fast and Robust Attention-Free Heterogeneous Graph Convolutional Network

Low cost Non freeze-drying construction strategy of cellulose aerogels for efficient solar desalination

Aerogel-like materials with three-dimensional pore structures show great potential in the field of efficient solar-driven interfacial evaporation by virtue of their fast water transport as well as efficient energy confinement properties. However, the preparation of aerogels is currently limited by high-cost strategies such as freeze drying or supercritical drying. Here, we present a foam template-assisted (FTA) strategy for atmospheric pressure drying of cellulose nanofiber (CNF) aerogels. Building a robust CNF network is the key to resisting surface tension during water evaporation and achieving drying in ambient environments, which is achieved through cleverly designed ionic cross-linking with the help of FTA. Cellulose aerogel (CA) constructed using the FTA strategy exhibit lightweight (9 mg cm−3) and porous properties (porosity: > 99 %). The tannic-coated CA (TCA) was achieved by further polymerizing tannic acid on the CA surface, which has excellent photothermal conversion capability. Meanwhile, TCA achieved an impressive evaporation rate of 3.58 kg m-2h−1 and an energy efficiency of 97.2 % at 1sun radiation by virtue of its excellent energy confinement and water management capabilities. This strategy provides a promising solution for the low-cost construction of aerogels as well as for the realization of efficient solar-driven interfacial evaporators.

Remote sensing image classification based on non-linear enhanced attention mechanism

Abstract In recent years, Transformer technology has gradually shown promising applications in the field of computer vision, becoming a research hotspot. However, traditional vision Transformers suffer from significant computational burdens. Although previous studies have attempted to alleviate this issue by reducing the computational load of attention mechanisms, their methods still require improvement. Additionally, past research often overlooked the non-linear representation of values within attention mechanisms, posing another challenge to address. Therefore, this paper proposes a novel attention mechanism aimed at reducing the computational burden of traditional attention mechanisms while enhancing their ability to express non-linearities in values. Furthermore, conventional feedforward neural networks typically output results directly after passing through an activation function, which may limit the model’s representational capacity due to the simplistic structure leading to singular feature information. To address this issue, this study introduces a more robust feature fusion convolutional feedforward network. Through these methods, the aim is to enhance the accuracy of Transformers in the field of remote-sensing image classification.

Massive experimental quantification of amyloid nucleation allows interpretable deep learning of protein aggregation.

Protein aggregation is a pathological hallmark of more than fifty human diseases and a major problem for biotechnology. Methods have been proposed to predict aggregation from sequence, but these have been trained and evaluated on small and biased experimental datasets. Here we directly address this data shortage by experimentally quantifying the amyloid nucleation of >100,000 protein sequences. This unprecedented dataset reveals the limited performance of existing computational methods and allows us to train CANYA, a convolution-attention hybrid neural network that accurately predicts amyloid nucleation from sequence. We adapt genomic neural network interpretability analyses to reveal CANYA's decision-making process and learned grammar. Our results illustrate the power of massive experimental analysis of random sequence-spaces and provide an interpretable and robust neural network model to predict amyloid nucleation.

The Impact Of Female Entrepreneurs On Local Economies

This paper looks into the contribution of female entrepreneurs to regional economies and gives further detail on the contribution they may make to innovation, economic growth, and community development. It proves that women-owned businesses do not give variety to the market but are also significant in generating employment and reinvesting revenues into their local communities, which causes a trickle effect and increases well-being within the general economy. The paper will discuss case studies on successful and famous female entrepreneurs: Mindy Kaling, founder of Kaling International; Sara Blakely, founder of Spanx; Whitney Wolfe Herd, co-founder of Bumble; and Nadia Hussain, owner/creator of Nadia's Kitchen. The study of each woman will show precisely how these women have revolutionized industries and sparked local economies through their unique ideas and the leading force of this entrepreneurship. Success and stories of Sara Blakely with Spanx also prove that one entrepreneurial vision can change an entire industry in the blink of a second by creating a new market segment and thousands of jobs. Also, the creation of Bumble has disrupted not only social networking and dating apps but also created great economic benefits with regard to jobs and technology. Mindy Kaling’s impact through Kaling International showcases the potential for creative entrepreneurs to shape media and entertainment, influencing cultural narratives while supporting local talent and economic growth. Nadia Hussain’s Nadia’s Kitchen underscores the role of culinary entrepreneurship in promoting cultural diversity and community engagement, alongside creating new employment opportunities. The paper also explores the critical success factors contributing to the achievements of these female entrepreneurs, including access to funding, robust support networks, educational opportunities, and favorable policy environments. It highlights how these factors together help bring about entrepreneurship among women, succeeding in their barrier-breaking business enterprises with unique strengths. It therefore points to wider economic and social benefits of female entrepreneurship through in-depth examination of these factors. It goes further to argue for the necessity for continued and increased support for women in business through policies and initiatives that can further make their impact on local economies more solid. This, the research argues, is how investing in female entrepreneurs helps build economic development into more inclusive and resilient communities

Practical Approach to Longitudinal Neurologic Care of Adults With X-Linked Adrenoleukodystrophy and Adrenomyeloneuropathy.

Although X-linked adrenoleukodystrophy (ALD) has historically been considered a childhood disease managed by pediatric neurologists, it is one of the most common leukodystrophies diagnosed in adulthood. An increase in both male and female adults reaching diagnosis due to familial cases identified by state newborn screening panels and more widespread use of genetic testing results in a large cohort of presymptomatic or early symptomatic adults. This population is in urgent need of standardized assessments and follow-up care. Adults with ALD/adrenomyeloneuropathy (AMN) may be diagnosed in a variety of ways, including after another family member is identified via genetic testing or newborn screening, presenting for symptomatic evaluation, or following diagnosis with primary adrenal insufficiency. Significant provider, patient, and systems-based barriers prevent adult patients with ALD/AMN from receiving appropriate care, including lack of awareness of the importance of longitudinal neurologic management. Confirmation of and education about the diagnosis should be coordinated in conjunction with a genetic counselor. Routine surveillance for adrenal insufficiency and onset of cerebral ALD (CALD) in men should be performed systematically to avoid preventable morbidity and mortality. While women with ALD do not usually develop cerebral demyelination or adrenal insufficiency, they remain at risk for myeloneuropathy and are no longer considered "carriers." After diagnosis, patients should be connected to the robust support networks, foundations, and research organizations available for ALD/AMN. Core principles of neurologic symptom management parallel those for patients with other etiologies of progressive spastic paraplegia. Appropriate patient candidates for hematopoietic stem cell transplant (HSCT) and other investigational disease-modifying strategies require early identification to achieve optimal outcomes. All patients with ALD/AMN, regardless of sex, age, or symptom severity, benefit from a multidisciplinary approach to longitudinal care spearheaded by the neurologist. This review proposes key strategies for diagnostic confirmation, laboratory and imaging surveillance, approach to symptom management, and guidance for identification of appropriate candidates for HSCT and investigational treatments.

Introducing explainable artificial intelligence to property prediction in metal additive manufacturing

Industry demands not only accurate predictability, but also the ability to explain predictions made by machine learning models. In the same endeavor, this work introduces eXplainable Artificial Intelligence (XAI) to mechanical property prediction of additively manufactured materials. Historical data for as-built Ti-6Al-4 V alloy manufactured via selective laser melting (SLM) was mined to generate a comprehensive dataset. Robust Gaussian Process Regression (GPR) and Neural Network (NN) were built using a hefty 189 training data arguments. Alongside primary SLM process parameters, novel features such as sample porosity and build direction, which are known to have direct impact on strength, were also utilized. The optimized GPR model exhibited mean absolute errors of 23.9 MPa and 0.58 % when inferencing on test tensile strength and elongation, respectively. On the other hand, the optimized NN model performed slightly worse with errors of 28.24 MPa and 0.97 % for respective tensile properties. Beyond training and testing of the models, they were tested for explainability against different core levels of human-centric understanding put forth by XAI community. It was ultimately concluded that explainability may come at the cost of accuracy.

A novel colorless flexible polymer aerogel with stable amine linkage enabled superior formaldehyde removal and multifunctional application

Massive formaldehyde emissions have seriously endangered environment and human health, while predictably polymer aerogel is a promising candidate for formaldehyde (FA) removal, whereas low FA adsorption capacity, high volume shrinkage, and mechanical brittleness restrict its further development. Herein, based on “rigid-soft combination” polycondensation strategy, trifunctional 1,3,5-benzenetricarboxaldehyde (BTC) and polyetheramine Jeffamine T403 are selected as building blocks to construct a novel amine-linked polymer aerogel (rBTA) with hierarchical micro-nano porous structure via mild sol-gel method and freeze-drying process. Stable amine linkages in rBTA are formed by NaBH4 reduction of imine bonds, improving acid/alkali resistance and chemical stability greatly, and creating numerous active amine sites for reaction with FA. Formation of robust 3D covalent network facilitates to simultaneously achieve ultra-low-shrinkage (5.4%), high specific compressive modulus (5.32 kN·m·kg-1), remarkable mechanical durability/fatigue-resistance, flexibility, and machinability. Benefiting from synergistic physical-chemical adsorption effect, FA adsorption capacity and removal ratio of rBTA reach as high as 459.84 mg·g-1 and 97.87%, respectively. Colorless rBTA is further ground into powders for preparing polyoxymethylene (POM)/rBTA composite without affecting its color and formaldehyde emission amount at 60/230°C effectively decreases by 66.87%/79.55%. Furthermore, rBTA exhibits hydrophobicity, low thermal conductivity, and high solar reflectance, being successfully applied in emulsion separation, thermal insulation, and energy-efficient buildings fields.

Urban flooding damage prediction in matrix scenarios of extreme rainfall using a convolutional neural network

The pivotal role of quantitative risk assessment in managing urban stormwater is underscored by the high spatial heterogeneity and complex non-stationarity complex of urban flooding. Conventional methods, reliant on measurable data, often fall short in accurately mapping spatial variations and gauging the full impacts of urban flooding. Addressing this gap, this study proposed a robust convolutional neural networks (CNN)-based tool, specifically designed for urban flooding risk and economic losses estimation under extreme design rainfall scenarios. Furthermore, the study validates the transferability of CNN models trained in data-abundant regions to similar but data-scarce regions, using Guangzhou as a case study for urban flooding damage prediction. The findings reveal that the most affected areas, particularly the old, densely built-up urban areas in the south-central Guangzhou, are susceptible to significant economic losses during extreme rainfall events. Notably, under the most severe scenario (Scenario 5), estimated economic losses amount to approximately $6359.91 million, with industrial and residential sectors bearing the brunt, accounting for 28.29 % and 39.94 % of the total losses, respectively. These insights are crucial for prioritizing mitigation efforts and formulating effective evacuation strategies in high-risk areas, ultimately aiding in the reduction of economic losses.

FlightTrackAI: a robust convolutional neural network-based tool for tracking the flight behaviour of Aedes aegypti mosquitoes.

Monitoring the flight behaviour of mosquitoes is crucial for assessing their fitness levels and understanding their potential role in disease transmission. Existing methods for tracking mosquito flight behaviour are challenging to implement in laboratory environments, and they also struggle with identity tracking, particularly during occlusions. Here, we introduce FlightTrackAI, a robust convolutional neural network (CNN)-based tool for automatic mosquito flight tracking. FlightTrackAI employs CNN, a multi-object tracking algorithm, and interpolation to track flight behaviour. It automatically processes each video in the input folder without supervision and generates tracked videos with mosquito positions across the frames and trajectory graphs before and after interpolation. FlightTrackAI does not require a sophisticated setup to capture videos; it can perform excellently with videos recorded using standard laboratory cages. FlightTrackAI also offers filtering capabilities to eliminate short-lived objects such as reflections. Validation of FlightTrackAI demonstrated its excellent performance with an average accuracy of 99.9%. The percentage of correctly assigned identities after occlusions exceeded 91%. The data produced by FlightTrackAI can facilitate analysis of various flight-related behaviours, including flight distance and volume coverage during flights. This advancement can help to enhance our understanding of mosquito ecology and behaviour, thereby informing targeted strategies for vector control.

Guest Editors’ Introduction: Special Issue on Robust and Resilient Future Communication Networks

Guest Editors’ Introduction: Special Issue on Robust and Resilient Future Communication Networks

Biochanin-A co-crystal formulation improves bioavailability and ameliorates cerulein-induced pancreatitis by attenuating the inflammation

Co-crystallization of a therapeutic ingredient with an appropriate co-former is a powerful technique to augment the physicochemical and pharmacokinetic properties and the effectiveness of Active Pharmaceutical Ingredients (APIs). Biochanin A (BCA), a flavonoid with medicinal potential, is limited by poor solubility and low oral bioavailability. This study aimed to design and develop a novel BCA-nicotinamide cocrystal as BCC to enhance BCA’s oral bioavailability and explore its therapeutic potential for ameliorating cerulein-induced acute pancreatitis (CIAP) by elucidating the target identification utilizing tissue/serum metabolite profiles. The cocrystal was designed by the supramolecular synthon approach and characterized by single-crystal X-ray diffraction that confirms a robust three-dimensional hydrogen-bonded network of BCA and Nicotinamide (NCT) in the crystal. FT-IR and DSC were used to analyze the cocrystal’s intermolecular interactions and thermal behavior. BCC exhibited enhanced solubility and drug release compared to BCA alone, resulting in enhanced oral bioavailability and pancreatic tissue concentration. Comparing BCC to BCA in the CIAP model, BCC therapy remarkably reduced cerulein-induced pancreatitis, evidenced by significant reductions in inflammation, acinar cell atrophy, and amylase levels in pancreatic tissues. Further, the cocrystal formulation also down-regulated the oxidative stress markers, inflammatory cytokines and macrophage-related proteins. The study has identified distinct metabolomic signatures linked with AP with the help of Orbitrap Exploris mass spectrometry, which could pave the way for creating focused diagnostic tools for a better prognosis. In conclusion, these results offer new insights into exploring mechanistic pathways associated with specific biomarkers and underscore BCC cocrystals as a promising approach to enhance BCA’s therapeutic potential.

Expansion of d-spacing of boehmite for enhanced phosphate adsorption via hydrogen bond network

The optimization of the interlayer water content to increase the d-spacing of layered materials has recently attracted interest in pollutant removal. Several studies have explored phosphate adsorption onto boehmite considering its high surface area; however, the influence of its d-spacing on the adsorption performance remains unexplored. Therefore, our study aimed to investigate the impact of the synthesis pH of boehmite on the d-spacing and adsorption performance of phosphate ions. For this purpose, boehmite samples were synthesized under various pH conditions at 100°C to be tested for phosphate adsorption. As a result of this mild synthesis condition, the number of reactive hydroxyl groups in boehmite samples ranged from 7.02 – 9.02mmol/g, nearly four times higher than the reported boehmites. Boehmite prepared under acidic conditions exhibits a relatively high interlayer water content and an enlarged d-spacing from 0.640 to 0.996nm, resulting in superior phosphate adsorption (215mg/g) within one minute. We also confirmed that phosphate adsorption increased the binding energy states of aluminum and oxygen in boehmite, indicating a robust hydrogen bond network with phosphate ions in the inner space. In contrast, the adsorption decreased the binding energy for other types of boehmite, indicating charge interaction. Competitive adsorption tests also showed the strong affinity of boehmite for phosphate ions, even in the presence of fluoride ions. These findings highlight the importance of the interlayer water molecules in controlling the d-spacing of boehmite, indicating their critical role in removing anions from water.

Targeting attack activity-driven networks.

Real-world complex systems demonstrated temporal features, i.e., the network topology varies with time and should be described as temporal networks since the traditional static networks cannot accurately characterize. To describe the deliberate attack events in the temporal networks, we propose an activity-based targeted attack on the activity-driven network to investigate temporal networks' temporal percolation properties and resilience. Based on the node activity and network mapping framework, the giant component and temporal percolation threshold are solved according to percolation theory and generating function. The theoretical results coincide with the simulation results near the thresholds. We find that targeted attacks can affect the temporal network, while random attacks cannot. As the probability of a highly active node being deleted increases, the temporal percolation threshold increases, and the giant component increases, thus enhancing robustness. When the network's activity distribution is extremely heterogeneous, network robustness decreases consequently. These findings help us to analyze and understand real-world temporal networks.