检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Huang, Zhongzheng Wu, Jiawei Wang, Tao Li, Zuoyong Ioannou, Anastasia College of Computer and Control Engineering Minjiang University China College of Computer and Data Science Fuzhou University China College of Electrical and Mechanical Engineering Fujian Agriculture and Forestry University China International Digital Economy College Minjiang University China Department of Computer Science and Engineering European University Cyprus Cyprus Cyprus

Despite the success of deep neural networks in medical image classification, the problem remains challenging as data annotation is time-consuming, and the class distribution is imbalanced due to the relative scarcity of diseases. To address this problem, we propose Class-Specific Distribution Alignment (CSDA), a semi-supervised learning framework based on self-training that is suitable to learn from highly imbalanced datasets. Specifically, we first provide a new perspective to distribution alignment by considering the process as a change of basis in the vector space spanned by marginal predictions, and then derive CSDA to capture class-dependent marginal predictions on both labeled and unlabeled data, in order to avoid the bias towards majority classes. Furthermore, we propose a Variable Condition Queue (VCQ) module to maintain a proportionately balanced number of unlabeled samples for each class. Experiments on three public datasets HAM10000, CheXpert and Kvasir show that our method provides competitive performance on semi-supervised skin disease, thoracic disease, and endoscopic image classification tasks. Copyright © 2023, The Authors. All rights reserved.

关键词： Alignment

来源：评论

学校读者我要写书评

暂无评论

Transferable Physical Adversarial Patch Attack for Remote Sensing Object Detection

Transferable Physical Adversarial Patch Attack for Remote Se...

引用

IEEE International Symposium on Geoscience and Remote Sensing (IGARSS)

作者： Di Wang Wenxuan Zhu Ke Li Xiao Gao Pengfei Yang School of Computer Science and Technology Xidian University Xi’an China College of Information and Control Engineering Xi’an University of Architecture and Technology Xi’an China

ISBN: (数字)9798350360325

ISBN: (纸本)9798350360332

Deep neural networks (DNNs) have been widely used in remote sensing but demonstrated to be vulnerable with adversarial examples. By adding elaborately designed perturbations on the clean images, DNNs may output wrong prediction. Research on adversarial attack contributes to the study of model robustness. However, previous methods mainly focus on white-box scenario or digital domain for classification tasks, while the vulnerability of remote sensing detectors has not been fully explored. Aiming at attacking black-box remote sensing detectors in physical domain, we propose to generate a transferable physical adversarial patch (TPAP) as the perturbations. Specifically, the initial patch is optimized by a U-Net and modified by the plane mask and position mask before applied to the clean image. By attacking a surrogate model, TPAP can be transferred to the target model. Abundant experimental results validate the attack ability of TPAP and evaluate the robustness of current one-stage detectors.

关键词： YOLO Perturbation methods Closed box Detectors Robustness Sensors Task analysis

来源：评论

学校读者我要写书评

暂无评论

Predicting Off-Design Performance of Axial Compressors Using Multilayer Neural Network Model for Surge control

SSRN

引用

SSRN 2024年

作者： Elmelegy, Mohamed Ali Elhosseini, Mostafa A. Ali, Hesham Arafat Haikal, Amira Y. DLNG Damietta LNG Egypt College of Computer Science and Engineering Taibah University Yanbu Saudi Arabia Department of Computers and Control Systems Engineering Faculty of Engineering Mansoura University Mansoura Egypt

Incidence of surge within axial compressors profoundly influences efficacy, reliability of aero-engines. Conventional methodologies in engine design have predominantly concentrated on precise ,efficient forecasting critical characteristics during such occurrences. paper presents novel approach for predicting surge phenomenon in axial compressors. Surge is a major issue in operation axial compressors, causing reduced performance, increased wear, tear, complete machine failure. proposed approach utilizes NN model based on compressor's design data to forecast behavior of compressor's working map under off-design conditions. model is verified using real data collected over five years during compressor operation, including opening angle (IGV), important parameter often ignored or assumed to be fixed in previous studies. results show NN-based approach effectively predicts compressor's behavior under different operating conditions, providing valuable insights into onset of surge , offering real-time surge prediction ,control ,ability to predict compressor's working map enables optimizing its performance , controlling instability. model was validated by comparing its predictions with real operating conditions observed over five years across various compressor operating modes. As result, predictive model enables safe operation of axial compressors by mitigating instability risks during operation. study's findings highlight model's effectiveness, as evidenced by minimal error observed between predicted and actual data collected across various compressor operating modes. model exhibits strong performance, characterized by high regression factor of 0.92667 , low mean squared error of 3.04465 x 10^-3. numerical values underscore reliability , precision proposed approach in predicting operation of axial compressor under diverse conditions, ultimately contributing to enhanced stability , safety during compressor operations. © 2024, The Authors. All rights reserved.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Interval Observer-Based Attack Detection in Smart Grid Under FlexRay Communication Protocols

Interval Observer-Based Attack Detection in Smart Grid Under...

引用

China International Electrical and Energy Conference (CIEEC)

作者： Man Hu Yong Zhang Mengmeng Su Mingfei Ban North China Branch Of State Grid Corporation Of China Beijing China Northeast Forestry University College of Computer and Control Engineering Harbin China

ISBN: (数字)9798350359558

ISBN: (纸本)9798350359565

In this paper, the detection of false data injection attacks (FDIA) in the smart grid based on interval observer is studied under the FlexRay communication protocol. The FDIA is highly stealthy and can modify the sensor data to affect system performance without being detected by common detection mechanisms. A cyber-attack detection mechanism against FDIA is proposed in this paper. Firstly, the smart grid is divided into several sub-areas for analysis and modeling, and a novel FlexRay communication protocol is designed to schedule the transmission of information in each sub-area. Next, an interval observer considering internal state boundaries and disturbances is designed based on the information after transmission to estimate the upper and lower bounds of the system states. An algorithm for attack detection against FDIA using interval residuals from interval observers is proposed. In addition, using the augmentation technique and the Lyapunov stability analysis method, sufficient conditions to ensure the stability of the estimation error system are given in the form of linear matrix inequality (LMI), and the interval observer gains are obtained by solving the LMI. Finally, the correctness and effectiveness of the proposed method is verified by a simulation example.

关键词： Sufficient conditions Schedules Protocols System performance Observers Power system stability Stability analysis

来源：评论

学校读者我要写书评

暂无评论

Distributed Stochastic Bilevel Optimization: Improved Complexity and Heterogeneity Analysis

arXiv

引用

arXiv 2023年

作者： Niu, Youcheng Xu, Jinming Sun, Ying Huang, Yan Chai, Li College of Control Science and Engineering Zhejiang University China School of Electrical Engineering and Computer Science The Pennsylvania State University United States

This paper consider solving a class of nonconvex-strongly-convex distributed stochastic bilevel optimization (DSBO) problems with personalized inner-level objectives. Most existing algorithms require computational loops for hypergradient estimation, leading to computational inefficiency. Moreover, the impact of data heterogeneity on convergence in bilevel problems is not explicitly characterized yet. To address these issues, we propose LoPA, a loopless personalized distributed algorithm that leverages a tracking mechanism for iterative approximation of inner-level solutions and Hessian-inverse matrices without relying on extra computation loops. Our theoretical analysis explicitly characterizes the heterogeneity across nodes denoted by b, and establishes a b23 sublinear rate of O((1−ρ1)2K + (1−ρ) 32 K 23 + √1K(σp + √1mσc)) without the boundedness of local hypergradients, where σp and σc represent the gradient sampling variances associated with the inner- and outer-level variables, respectively. We also develop a variant of LoPA based on gradient tracking to eliminate the impact of data heterogeneity, yielding an improved rate of O((1−ρ1)4K + √1K(σp + √1mσc)). The computational complexity of LoPA is of O(ϵ−2) to an ϵ-stationary point, matching the communication complexity due to the loopless structure, which outperforms existing counterparts for DSBO. Numerical experiments validate the effectiveness of the proposed algorithm. © 2023, CC BY.

关键词： Computational complexity

来源：评论

学校读者我要写书评

暂无评论

Lightweight defect detection method of punched nickel-plated steel strip based on GhostNet

Lightweight defect detection method of punched nickel-plated...

引用

2023 International Conference on Pattern Recognition, Machine Vision and Intelligent Algorithms, PRMVIA 2023

作者： Li, Jianqi Liang, Yincong Du, Rui Wan, Jingying Cao, Binfang Liu, Huiting Hunan University of Arts and Sciences Hunan Provincial Key Laboratory for Control Technology of Distributed Electric Propulsion Aircraft Changde China Jishou University Hunan Provincial Key Laboratory for Control Technology of Distributed Electric Propulsion Aircraft Changde China Hunan University of Arts and Sciences College of Computer and Electrical Engineering Changde China

ISBN: (纸本)9798350346596

Aiming at the problem that the defects generated in the production and transportation of punched nickel-plated steel strips are not easy to be detected by deep learning methods, a lightweight, low-redundancy, and high-precision detection method is proposed in this paper. Firstly, a feature extraction network based on GhostNet is constructed, which reduces the amount of computation and feature redundancy while ensuring accuracy. Then the ECA module is applied to the detection head to perform weighted fusion of the features of different channels for better differentiation. Finally, the YOLO detection head is used for multi-scale detection. In the experiment, the mAP of 84.86% was obtained by this method, which proves that this method can be applied to the actual steel strip defect: detection. © 2023 IEEE.

关键词： Redundancy

来源：评论

学校读者我要写书评

暂无评论

A Lightweight Multi-domain Multi-attention Progressive Network for Single Image Deraining 24

A Lightweight Multi-domain Multi-attention Progressive Netwo...

引用

32nd ACM International Conference on Multimedia, MM 2024

作者： Zhong, Junliu Li, Zhiyi Xiang, Dan Han, Maotang Li, Changsheng Gan, Yanfen School of Computer Science Guangzhou Maritime University Guangdong Guangzhou China Network and Information Service Center Guangdong University of Foreign Studies South China Business College Guangdong Guangzhou China School of Low Altitude Equipment and Intelligent Control Guangzhou Maritime University Guangdong Guangzhou China School of Computer and Information Engineering Guangdong Polytechnic of Industry and Commerce Guangdong Guangzhou China

ISBN: (纸本)9798400706868

Currently, the information processing in a spatial domain alone has intrinsic limitations that hinder the deep network's effectiveness (performance) improvement in a single image deraining. Moreover, the deraining networks' structures and learning processes are becoming increasingly intricate, leading to challenges in structural lightweight, and training and testing efficiency. We propose a lightweight multi-domain multi-attention progressive network (M2PN) to handle these challenges. For performance improvement, the M2PN backbone applies a simple progressive CNN-based structure consisting of the S same recursive M2PN modules. This recursive backbone with a skip connection mechanism allows for better gradient flow and helps to effectively capture low-to-high-level/scales spatial features in progressive structure to improve contextual information acquisition. To further complement acquired spatial information for better deraining, we conduct spectral analysis on the frequency energy distribution of rain steaks, and theoretically present the relationship between the spectral bandwidths and the unique falling characteristics and special morphology of rain steaks. We present the frequency-channel attention (FcA) mechanism and the spatial-channel attention (ScA) mechanism to fuse frequency-channel features and spatial features better to distinguish and remove rain steaks. The simple recursive network structure and effective multi-domain multi-attention mechanism serve as the M2PN to achieve superior performance and facilitate fast convergence during training. Furthermore, the M2PN structure, with a small network component quantity, shallow network channels, and few convolutional kernels, requires only 168K parameters, which is 1 to 2 orders of magnitude lower than the existing SOTA networks. The experimental results demonstrate that even with such a few network parameters, M2PN still achieves the best overall performance. © 2024 ACM.

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

Mixed-Integer Optimal control With Disjunctive Characteristic Time Constraints for Energy-Optimal Speed Planning

引用

IEEE Transactions on Transportation Electrification 2025年

作者： Dong, Shiying Chen, Hong Luo, Xi Meng, Xiangyu Huang, Yanjun Liu, Qifang Gao, Bingzhao Jilin University Department of Control Science and Engineering Changchun China Tongji University College of Automotive Studies Shanghai China Louisiana State University Division of Electrical and Computer Engineering Baton RougeLA70803 United States

This paper considers the energy optimization problem with multiple time window constraints, which is motivated by the eco-driving of connected and automated vehicles (CAVs) in urban traffic. Timely crossing of signalized intersections defines complex spatio-temporal constraints, and multiple available open time windows introduce the discontinuous feasible domains of crossing times. The CAVs should select an open time-window for each intersection, determine the crossing times, and optimize the speed profiles accordingly, which turns out to be a hard-to-solve optimization problem. To address this challenge, this paper proposes a unifying approach that transforms the energy optimization problem into a novel mixed-integer optimal control problem with characteristic time constraints (MIOCP-CT). This approach facilitates the simultaneous optimization of open time windows, crossing times, and speed profiles within a single optimization problem, which can be solved by our proposed novel numerical solution procedure. The application of the proposed method in the eco-driving of CAVs approaching multiple signalized intersections demonstrates its performance within real-world urban driving cycles. The experimental results of the hardware-in-the-loop (HiL) demonstrate a superior energy-saving performance of 5.83% compared to existing methods, while also achieving real-time feasibility within 100 milliseconds. © 2015 IEEE.

关键词： Mixed-integer linear programming

来源：评论

学校读者我要写书评

暂无评论

An Enhanced Analysisof Traffic Intelligence in Smart Cities Using Sustainable Deep Radial Function

arXiv

引用

arXiv 2024年

作者： Ismaeel, Ayad Ghany Mary, S.J. Jereesha Anitha, C. Logeshwaran, Jaganathan Mahmood, Sarmad Nozad Alani, Sameer Shather, Akram H. Computer Technology Engineering College of Engineering Technology Al-Kitab University Kirkuk36001 Iraq Annai Velankanni College of Engineering Potalkulam Kanyakumari629401 India Saveetha School of Engineering Chennai600124 India Department of Electronics and Communication Engineering Sri Eshwar College of Engineering Coimbatore641202 India Electronic and Control Engineering Techniques Technical Engineering College Northern Technical University Kirkuk Iraq Computer Center University of Anbar Baghdad55431 Iraq Department of Computer Engineering Technology Al-Kitab University Altun Kopru Kirkuk Iraq

Smart cities have revolutionized urban living by incorporating sophisticated technologies to optimize various aspects of urban infrastructure, such as transportation systems. Effective traffic management is a crucial component of smart cities, as it has a direct impact on the quality of life of residents and tourists. Utilizing deep radial basis function (RBF) networks, this paper describes a novel strategy for enhancing traffic intelligence in smart cities. Traditional methods of traffic analysis frequently rely on simplistic models that are incapable of capturing the intricate patterns and dynamics of urban traffic systems. Deep learning techniques, such as deep RBF networks, have the potential to extract valuable insights from traffic data and enable more precise predictions and decisions. In this paper, we propose an RBF-based method for enhancing smart city traffic intelligence. Deep RBF networks combine the adaptability and generalization capabilities of deep learning with the discriminative capability of radial basis functions. The proposed method can effectively learn intricate relationships and nonlinear patterns in traffic data by leveraging the hierarchical structure of deep neural networks. The deep RBF model can learn to predict traffic conditions, identify congestion patterns, and make informed recommendations for optimizing traffic management strategies by incorporating theserich and diverse data. To evaluate the efficacy of our proposed method, extensive experiments and comparisons with real-world traffic datasets from a smart city environment were conducted. In terms of prediction accuracy and efficiency, the results demonstrate that the deep RBF-based approach outperforms conventional traffic analysis methods. Smart city traffic intelligence is enhanced by the model capacity to capture nonlinear relationships and manage large-scale data sets. © 2024, CC BY-NC-ND.

关键词： Radial basis function networks

来源：评论

学校读者我要写书评

暂无评论

LawLuo: A Multi-Agent Collaborative Framework for Multi-Round Chinese Legal Consultation

arXiv

引用

arXiv 2024年

作者： Sun, Jingyun Dai, Chengxiao Luo, Zhongze Chang, Yangbo Li, Yang College of Computer and Control Engineering Northeast Forestry University Faculty of Information and Communication Technology Universiti Tunku Abdul Rahman College of Architecture and Arts Design Shijiazhuang Tiedao University

Legal Large Language Models (LLMs) have shown promise in providing legal consultations to non-experts. However, most existing Chinese legal consultation models are based on single-agent systems, which differ from real-world legal consultations, where multiple professionals collaborate to offer more tailored responses. To better simulate real consultations, we propose LawLuo, a multi-agent framework for multi-turn Chinese legal consultations. LawLuo includes four agents: the receptionist agent, which assesses user intent and selects a lawyer agent;the lawyer agent, which interacts with the user;the secretary agent, which organizes conversation records and generates consultation reports;and the boss agent, which evaluates the performance of the lawyer and secretary agents to ensure optimal results. These agents’ interactions mimic the operations of real law firms. To train them to follow different legal instructions, we developed distinct fine-tuning datasets. We also introduce a case graph-based RAG to help the lawyer agent address vague user inputs. Experimental results show that LawLuo outperforms baselines in generating more personalized and professional responses, handling ambiguous queries, and following legal instructions in multi-turn conversations. Our full code and constructed datasets will be open-sourced upon paper acceptance. © 2024, CC BY.

关键词： Modeling languages

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：