Search

Your search keyword '"Wang, Longwei"' showing total 237 results
Start Over Author "Wang, Longwei"
237 results on '"Wang, Longwei"'

1. Enhancing Adversarial Robustness of Deep Neural Networks Through Supervised Contrastive Learning

Author

Wang, Longwei, Nayyem, Navid, and Rakin, Abdullah

Subjects
Computer Science - Machine Learning, Computer Science - Artificial Intelligence
Abstract

Adversarial attacks exploit the vulnerabilities of convolutional neural networks by introducing imperceptible perturbations that lead to misclassifications, exposing weaknesses in feature representations and decision boundaries. This paper presents a novel framework combining supervised contrastive learning and margin-based contrastive loss to enhance adversarial robustness. Supervised contrastive learning improves the structure of the feature space by clustering embeddings of samples within the same class and separating those from different classes. Margin-based contrastive loss, inspired by support vector machines, enforces explicit constraints to create robust decision boundaries with well-defined margins. Experiments on the CIFAR-100 dataset with a ResNet-18 backbone demonstrate robustness performance improvements in adversarial accuracy under Fast Gradient Sign Method attacks., Comment: 8 pages, 11 figures

Published
2024

2. Bridging Interpretability and Robustness Using LIME-Guided Model Refinement

Author

Nayyem, Navid, Rakin, Abdullah, and Wang, Longwei

Subjects
Computer Science - Machine Learning, Computer Science - Artificial Intelligence
Abstract

This paper explores the intricate relationship between interpretability and robustness in deep learning models. Despite their remarkable performance across various tasks, deep learning models often exhibit critical vulnerabilities, including susceptibility to adversarial attacks, over-reliance on spurious correlations, and a lack of transparency in their decision-making processes. To address these limitations, we propose a novel framework that leverages Local Interpretable Model-Agnostic Explanations (LIME) to systematically enhance model robustness. By identifying and mitigating the influence of irrelevant or misleading features, our approach iteratively refines the model, penalizing reliance on these features during training. Empirical evaluations on multiple benchmark datasets demonstrate that LIME-guided refinement not only improves interpretability but also significantly enhances resistance to adversarial perturbations and generalization to out-of-distribution data., Comment: 10 pages, 15 figures

Published
2024

3. Beyond Idle Channels: Unlocking Idle Space with Signal Alignment in Massive MIMO Cognitive Radio Networks

Author

Zhu, Weidong, Li, Xueqian, Wang, Longwei, and Zhang, Zheng

Subjects
Electrical Engineering and Systems Science - Signal Processing
Abstract

Cognitive radio networks (CRNs) have traditionally focused on utilizing idle channels to enhance spectrum efficiency. However, as wireless networks grow denser, channel-centric strategies face increasing limitations. This paper introduces a paradigm shift by exploring the underutilized potential of idle spatial dimensions, termed idle space, in co-channel transmissions. By integrating massive multiple-input multiple-output (MIMO) systems with signal alignment techniques, we enable secondary users to transmit without causing interference to primary users by aligning their signals within the null spaces of primary receivers. We propose a comprehensive framework that synergizes spatial spectrum sensing, signal alignment, and resource allocation, specifically designed for secondary users in CRNs. Theoretical analyses and extensive simulations validate the framework, demonstrating substantial gains in spectrum efficiency, throughput, and interference mitigation. The results show that the proposed approach not only ensures interference-free coexistence with primary users but also unlocks untapped spatial resources for secondary transmissions., Comment: 9 pages, 7 figures

Published
2024

4. Dense Cross-Connected Ensemble Convolutional Neural Networks for Enhanced Model Robustness

Author

Wang, Longwei, Li, Xueqian, and Zhang, Zheng

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Artificial Intelligence
Abstract

The resilience of convolutional neural networks against input variations and adversarial attacks remains a significant challenge in image recognition tasks. Motivated by the need for more robust and reliable image recognition systems, we propose the Dense Cross-Connected Ensemble Convolutional Neural Network (DCC-ECNN). This novel architecture integrates the dense connectivity principle of DenseNet with the ensemble learning strategy, incorporating intermediate cross-connections between different DenseNet paths to facilitate extensive feature sharing and integration. The DCC-ECNN architecture leverages DenseNet's efficient parameter usage and depth while benefiting from the robustness of ensemble learning, ensuring a richer and more resilient feature representation., Comment: 6 pages, 1 figure

Published
2024

6. Numerical Simulation of Hydrodynamic Characteristics of Spur Dikes in Beijiang River and Its Influence on Navigation Safety

Author

WANG Longwei, LU Zhuoyin, ZHU Dantong, ZENG Jiaxing, CHENG Xiangju, and HUANG Dong

Subjects
spur dikes, navigation, hydrodynamic model, Beijiang River, River, lake, and water-supply engineering (General), TC401-506
Abstract

In order to deeply explore the influence of spur dikes in Beijiang River on navigation safety, this study adopted the calculation model of unstructured-grid finite volume coastal Dcean model (FVCOM), took the mainstream reach of 23 km of Beijiang River in Yingde section as the research area, constructed a three-dimensional hydrodynamic model, and verified the effectiveness of the model through measured data. On this basis, the hydrodynamic characteristics with and without spur dikes were simulated under the assumed constant inlet flow rate of 420 m³/s, and the abrupt bend section in the study area was analyzed. The calculation results show that the spur dike reduces the near-shore flow velocity by 44.6%~90.3%, increases the channel depth by 0~0.2 m, and elevates the mainstream flow velocity by 66.6%~121.8%. In addition, the spur dike converges the lateral mainstream. The degree of swing is reduced by up to 41.6%, and the amplitude of swing is reduced by up to 50.2%, which has a positive effect on the improvement of the navigability of the channel. However, the arrangement of the spur dike also increases some negative effects. For example, the unilateral long spur dikes lead to the increase of the opposite near-shore flow velocity by 73.3%~129.0%, the decrease of the horizontal mainstream curvature of the channel from 784.2 m to 774.5 m, and the decrease of the bending radius of the hydrodynamic axis by 14.0%~36.7%, which may constitute a potential threat to the navigation safety. The results of this study not only reveal the complex role of spur dikes in the hydrodynamic characteristics of a specific river section but also provide an important reference for the analysis of spur dikes on navigation safety in natural channels.

Published
2024
Full Text
View/download PDF

7. Substation WSN Coverage Optimization Technology Based on Improved Dragonfly Algorithm

Author

Zhang, Donglei, Fu, Jianding, Gao, Hongjian, Wang, Longwei, Du, Fei, Angrisani, Leopoldo, Series Editor, Arteaga, Marco, Series Editor, Chakraborty, Samarjit, Series Editor, Chen, Jiming, Series Editor, Chen, Shanben, Series Editor, Chen, Tan Kay, Series Editor, Dillmann, Rüdiger, Series Editor, Duan, Haibin, Series Editor, Ferrari, Gianluigi, Series Editor, Ferre, Manuel, Series Editor, Jabbari, Faryar, Series Editor, Jia, Limin, Series Editor, Kacprzyk, Janusz, Series Editor, Khamis, Alaa, Series Editor, Kroeger, Torsten, Series Editor, Li, Yong, Series Editor, Liang, Qilian, Series Editor, Martín, Ferran, Series Editor, Ming, Tan Cher, Series Editor, Minker, Wolfgang, Series Editor, Misra, Pradeep, Series Editor, Mukhopadhyay, Subhas, Series Editor, Ning, Cun-Zheng, Series Editor, Nishida, Toyoaki, Series Editor, Oneto, Luca, Series Editor, Panigrahi, Bijaya Ketan, Series Editor, Pascucci, Federica, Series Editor, Qin, Yong, Series Editor, Seng, Gan Woon, Series Editor, Speidel, Joachim, Series Editor, Veiga, Germano, Series Editor, Wu, Haitao, Series Editor, Zamboni, Walter, Series Editor, Tan, Kay Chen, Series Editor, Cai, Chunwei, editor, Qu, Xiaohui, editor, Mai, Ruikun, editor, Zhang, Pengcheng, editor, Chai, Wenping, editor, and Wu, Shuai, editor

Published
2024
Full Text
View/download PDF

10. Prediction of Corpse Model Drifting Trajectories in Pearl River Front Channel Based on Delft3D Model

Author

CHENG Xiangju, CHEN Zehai, LI Bin, LI Xu, BI Jinjin, and WANG Longwei

Subjects
Delft3D, tidal reach, drifting trajectory, prediction model, fitting, River, lake, and water-supply engineering (General), TC401-506
Abstract

To predict the drifting trajectories of bodies in the Pearl River Front Channel and assist local water police in locating the specific positions of drowning victims, this study uses Delft3D to construct a hydrodynamic model of the Pearl River Front Channel and conducts multiple on-site drifting tests using a corpse model. By fitting the surface flow velocity of the Pearl River Front Channel, the paper builds a prediction model for corpse model drifting with an R2 of 0.88. The results show that the corpse model released in the Pearl River Front Channel undergoes back-and-forth motion along the river channel under the influence of tides and flows, gradually drifting downstream. In the drifting model validation, the drifting speed and direction of the corpse model are generally consistent with the tidal current, and the error in simulation results is within 1 km, with a final distance error rate of less than 15%. In a case study, the simulated results for an adult female corpse have an error of approximately 300 m. External forces such as wave force caused by boats result in a north-south directional deviation in the drifting trajectory, indicating the need to further improve the simulation effect. The derivation mode of the model is also applicable to other tidal rivers, making it possible to measure the drifting trajectories of bodies and providing convenience and reference for body recovery work and police case processing.

Published
2024
Full Text
View/download PDF

16. Neurons Activation Visualization and Information Theoretic Analysis

Author

Wang, Longwei and Chen, Peijie

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Understanding the inner working mechanism of deep neural networks (DNNs) is essential and important for researchers to design and improve the performance of DNNs. In this work, the entropy analysis is leveraged to study the neurons activation behavior of the fully connected layers of DNNs. The entropy of the activation patterns of each layer can provide a performance metric for the evaluation of the network model accuracy. The study is conducted based on a well trained network model. The activation patterns of shallow and deep layers of the fully connected layers are analyzed by inputting the images of a single class. It is found that for the well trained deep neural networks model, the entropy of the neuron activation pattern is monotonically reduced with the depth of the layers. That is, the neuron activation patterns become more and more stable with the depth of the fully connected layers. The entropy pattern of the fully connected layers can also provide guidelines as to how many fully connected layers are needed to guarantee the accuracy of the model. The study in this work provides a new perspective on the analysis of DNN, which shows some interesting results., Comment: the paper is not so well written and need to be revised

Published
2019

27. Layer-Wise Entropy Analysis and Visualization of Neurons Activation

Author

Wang, Longwei, Chen, Peijie, Wang, Chengfei, Wang, Rui, Akan, Ozgur, Editorial Board Member, Bellavista, Paolo, Editorial Board Member, Cao, Jiannong, Editorial Board Member, Coulson, Geoffrey, Editorial Board Member, Dressler, Falko, Editorial Board Member, Ferrari, Domenico, Editorial Board Member, Gerla, Mario, Editorial Board Member, Kobayashi, Hisashi, Editorial Board Member, Palazzo, Sergio, Editorial Board Member, Sahni, Sartaj, Editorial Board Member, Shen, Xuemin (Sherman), Editorial Board Member, Stan, Mircea, Editorial Board Member, Jia, Xiaohua, Editorial Board Member, Zomaya, Albert Y., Editorial Board Member, Gao, Honghao, editor, Feng, Zhiyong, editor, Yu, Jun, editor, and Wu, Jun, editor

Published
2020
Full Text
View/download PDF

28. Deep Reinforcement Learning Based Computation Offloading for Mobility-Aware Edge Computing

Author

Shi, Minyan, Wang, Rui, Liu, Erwu, Xu, Zhixin, Wang, Longwei, Akan, Ozgur, Editorial Board Member, Bellavista, Paolo, Editorial Board Member, Cao, Jiannong, Editorial Board Member, Coulson, Geoffrey, Editorial Board Member, Dressler, Falko, Editorial Board Member, Ferrari, Domenico, Editorial Board Member, Gerla, Mario, Editorial Board Member, Kobayashi, Hisashi, Editorial Board Member, Palazzo, Sergio, Editorial Board Member, Sahni, Sartaj, Editorial Board Member, Shen, Xuemin (Sherman), Editorial Board Member, Stan, Mircea, Editorial Board Member, Jia, Xiaohua, Editorial Board Member, Zomaya, Albert Y., Editorial Board Member, Gao, Honghao, editor, Feng, Zhiyong, editor, Yu, Jun, editor, and Wu, Jun, editor

Published
2020
Full Text
View/download PDF

29. DFT-Spread Based PAPR Reduction of OFDM for Short Reach Communication Systems

Author

Li, Yupeng, Wang, Yaqi, Wang, Longwei, Angrisani, Leopoldo, Series Editor, Arteaga, Marco, Series Editor, Panigrahi, Bijaya Ketan, Series Editor, Chakraborty, Samarjit, Series Editor, Chen, Jiming, Series Editor, Chen, Shanben, Series Editor, Chen, Tan Kay, Series Editor, Dillmann, Rüdiger, Series Editor, Duan, Haibin, Series Editor, Ferrari, Gianluigi, Series Editor, Ferre, Manuel, Series Editor, Hirche, Sandra, Series Editor, Jabbari, Faryar, Series Editor, Jia, Limin, Series Editor, Kacprzyk, Janusz, Series Editor, Khamis, Alaa, Series Editor, Kroeger, Torsten, Series Editor, Liang, Qilian, Series Editor, Martín, Ferran, Series Editor, Ming, Tan Cher, Series Editor, Minker, Wolfgang, Series Editor, Misra, Pradeep, Series Editor, Möller, Sebastian, Series Editor, Mukhopadhyay, Subhas, Series Editor, Ning, Cun-Zheng, Series Editor, Nishida, Toyoaki, Series Editor, Pascucci, Federica, Series Editor, Qin, Yong, Series Editor, Seng, Gan Woon, Series Editor, Speidel, Joachim, Series Editor, Veiga, Germano, Series Editor, Wu, Haitao, Series Editor, Zhang, Junjie James, Series Editor, Wang, Wei, editor, Liu, Xin, editor, Na, Zhenyu, editor, Jia, Min, editor, and Zhang, Baoju, editor

Published
2020
Full Text
View/download PDF

34. Detection of Print Head Defect Based on Image Processing Technology

Author

Zhao, Zhanfeng, Liu, Duo, Wang, Longwei, Akan, Ozgur, Editorial Board Member, Bellavista, Paolo, Editorial Board Member, Cao, Jiannong, Editorial Board Member, Coulson, Geoffrey, Editorial Board Member, Dressler, Falko, Editorial Board Member, Ferrari, Domenico, Editorial Board Member, Gerla, Mario, Editorial Board Member, Kobayashi, Hisashi, Editorial Board Member, Palazzo, Sergio, Editorial Board Member, Sahni, Sartaj, Editorial Board Member, Shen, Xuemin (Sherman), Editorial Board Member, Stan, Mircea, Editorial Board Member, Xiaohua, Jia, Editorial Board Member, Zomaya, Albert Y., Editorial Board Member, Jia, Min, editor, Guo, Qing, editor, and Meng, Weixiao, editor

Published
2019
Full Text
View/download PDF

35. Feature Extraction and Identification of Pipeline Intrusion Based on Phase-Sensitive Optical Time Domain Reflectometer

Author

Zhao, Zhanfeng, Liu, Duo, Wang, Longwei, Liu, Shujun, Akan, Ozgur, Editorial Board Member, Bellavista, Paolo, Editorial Board Member, Cao, Jiannong, Editorial Board Member, Coulson, Geoffrey, Editorial Board Member, Dressler, Falko, Editorial Board Member, Ferrari, Domenico, Editorial Board Member, Gerla, Mario, Editorial Board Member, Kobayashi, Hisashi, Editorial Board Member, Palazzo, Sergio, Editorial Board Member, Sahni, Sartaj, Editorial Board Member, Shen, Xuemin (Sherman), Editorial Board Member, Stan, Mircea, Editorial Board Member, Xiaohua, Jia, Editorial Board Member, Zomaya, Albert Y., Editorial Board Member, Jia, Min, editor, Guo, Qing, editor, and Meng, Weixiao, editor

Published
2019
Full Text
View/download PDF

36. Nitrogen Vacancy Modulation of Tungsten Nitride Peroxidase-Mimetic Activity for Bacterial Infection Therapy

Author

Yang, Zhongwei, Wang, Longwei, Zhang, Xiaoyu, Zhang, Jian, Ren, Na, Ding, Longhua, Wang, Aizhu, Liu, Jing, Liu, Hong, and Yu, Xin

Abstract

Bacterial infections claim millions of lives every year, with the escalating menace of microbial antibiotic resistance compounding this global crisis. Nanozymes, poised as prospective substitutes for antibiotics, present a significant frontier in antibacterial therapy, yet their precise enzymatic origins remain elusive. With the continuous development of nanozymes, the applications of elemental N-modulated nanozymes have spanned multiple fields, including sensing and detection, infection therapy, cancer treatment, and pollutant degradation. The introduction of nitrogen into nanozymes not only broadens their application range but also holds significant importance for the design of catalysts in biomedical research. The synergistic interplay between W and N induces pivotal alterations in electronic configurations, endowing tungsten nitride (WN) with a peroxidase-like functionality. Furthermore, the introduction of N vacancies augments the nanozyme activity, thus amplifying the catalytic potential of WN nanostructures. Rigorous theoretical modeling and empirical validation corroborate the genesis of the enzyme activity. The meticulously engineered WN nanoflower architecture exhibits an exceptional ability in traversing bacterial surfaces, exerting potent bactericidal effects through direct physical interactions. Additionally, the topological intricacies of these nanostructures facilitate precise targeting of generated radicals on bacterial surfaces, culminating in exceptional bactericidal efficacy against both Gram-negative and Gram-positive bacterial strains along with notable inhibition of bacterial biofilm formation. Importantly, assessments using a skin infection model underscore the proficiency of WN nanoflowers in effectively clearing bacterial infections and fostering wound healing. This pioneering research illuminates the realm of pseudoenzyme activity and bacterial capture-killing strategies, promising a fertile ground for the development of innovative, high-performance artificial peroxidases.

Published
2024
Full Text
View/download PDF

38. Bimetallic Sulfides with Vacancy Modulation Exhibit Enhanced Electrochemical Performance.

Author

Guo, Jiawei, Zhao, Hongbo, Yang, Zhongwei, Wang, Longwei, Wang, Aizhu, Zhang, Jian, Ding, Longhua, Wang, Longfei, Liu, Hong, and Yu, Xin

Subjects
OXYGEN evolution reactions, CARBON fibers, ELECTRODE performance, ELECTROCHEMICAL electrodes, SULFIDES
Abstract

Transition bimetallic sulfides show significant promise for energy‐related applications because of their plentiful active sites and synergistic redox activity. However, limited pore size and low‐conductivity issues hinder their application. The structure of NiCo–S with rich sulfur vacancies is first predicted by density functional theory (DFT) calculations. Different sulfur vacancy concentrations are modeled by DFT calculations, and the results confirm that sulfur vacancies enhance the conductivity of the electrode material and are more beneficial for the adsorption of OH* species. It is verified by the differential charge density that the electric field formed on the surface of the electrode can lead to strong interfacial interactions by electron aggregation, which promotes electron/ion transfer kinetics. Furthermore, NiCo–S nanosheets are prepared on carbon cloth enriched with different concentrations of sulfur vacancies (denoted as NiCo‐Sv‐x, with x representing the concentration of sulfur vacancies) by sulfide etching NiCo‐MOF and annealing under H2/Ar atmosphere. The NiCo‐Sv‐x electrodes obtained are applied to the cathode of supercapacitors and the anode of the oxygen evolution reaction. Through combining experimental and theoretical analysis, the effect of vacancy defect engineering on the electrochemical performance of the electrode materials is further confirmed. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

50. The Environmental Adsorption and Combined Toxic Effects of Microplastics and Chemical Pollutants.

Author

WANG Longwei, ZHAO Jing, WANG Xinyu, and BAI Jianfeng

Abstract

Microplastics (MPs) are a kind of emerging pollutants, which have diversified sources, wide distribution, high abundance and great potential harm, and have been included in the key control list of emerging pollutants by Shanghai. MPs in water environment can adsorb heavy metals and organic pollutants, and the adsorption capacity is affected by MPs type, particle size, aging degree and environmental conditions. As a carrier, MPs can carry inorganic or organic pollutants into organisms, change the concentration and distribution of pollutants in organisms, and affect the toxic effects of pollutants. The type of combined action between MPs and chemical pollutants varies with the types of pollutants, and its specific mechanism remains to be further investigated. Therefore, this paper reviewed the latest development of domestic and foreign literatures, systematically summarized the sources and current status of MPs pollution in soil and freshwater environment in China, analyzed the adsorption behavior of MPs to chemical pollutants, the influencing factors of adsorption and the combined toxic effects of MPS on organisms, and prospected the future development direction of related researches. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results