检索结果-内蒙古大学图书馆

IEEE International Conference on Consumer Electronics (ICCE)

作者： Jedidah Mwaura Shunsuke Araki Muhammad Bisri Musthafa Samsul Huda Yasuyuki Nogami Graduate School of Computer Science and Systems Engineering Kyushu Institute of Technology Fukuoka Japan Faculty of Computer Science and Systems Engineering Kyushu Institute of Technology Fukuoka Japan Graduate School of Environmental Life Natural Science and Technology Okayama University Okayama Japan Green Innovation Center Okayama University Okayama Japan

ISBN: (数字)9798331521165

ISBN: (纸本)9798331521172

Currently, there are many ongoing attempts to incorporate transfer learning into Intrusion Detection systems. Transfer learning leverages on the knowledge of previous models thus reducing training time, improving accuracy and making it possible to train new models with limited data. However, it is difficult to incorporate such a technique into IoT devices with limited computational power. In this paper, we propose a Transfer Learning TinyML-based IDS framework with feature selection to be implemented on a microcontroller for detecting attacks on IoT devices. To implement the proposal, we leveraged knowledge achieved from our previous DDoS attack detection model to create a new transfer learning framework that could detect various IoT attacks such as malware attacks and zero-day attacks. To test the transfer learning framework, we trained it using two separate sets of datasets each containing different attacks. After each training, we evaluated the original model, converted it into a format compatible with a microcontroller and evaluated it further. The TL TinyML-based framework was evaluated based on its accuracy, model sizes and average inference times. As the results confirmed, the framework performed well and can successfully be implemented on a microcontroller without depleting its resources.

关键词： Training Accuracy Microcontrollers Computational modeling Transfer learning Intrusion detection Feature extraction Malware Data models Internet of Things

来源：评论

学校读者我要写书评

暂无评论

One-bit Compressed Sensing using Generative Models

arXiv

引用

arXiv 2025年

作者： Kafle, Swatantra Joseph, Geethu Varshney, Pramod K. ANDRO Computational Solutions NY United States Electrical Engineering and Computer Science Department Syracuse University NY United States Signal Processing Systems group Electrical Engineering Mathematics and Computer Science faculty the Delft University of Technology Netherlands

This paper addresses the classical problem of one-bit compressed sensing using a deep learning-based reconstruction algorithm that leverages a trained generative model to enhance the signal reconstruction performance. The generator, a pretrained neural network, learns to map from a low-dimensional latent space to a higher-dimensional set of sparse vectors. This generator is then used to reconstruct sparse vectors from their one-bit measurements by searching over its range. The presented algorithm provides an excellent reconstruction performance because the generative model can learn additional structural information about the signal beyond sparsity. Furthermore, we provide theoretical guarantees on the reconstruction accuracy and sample complexity of the algorithm. Through numerical experiments using three publicly available image datasets, MNIST, Fashion-MNIST, and Omniglot, we demonstrate the superior performance of the algorithm compared to other existing algorithms and show that our algorithm can recover both the amplitude and the direction of the signal from one-bit measurements. © 2025, CC BY.

关键词： Compressed sensing

来源：评论

学校读者我要写书评

暂无评论

RVAFM: Re-parameterizing Vertical Attention Fusion Module for Handwritten Paragraph Text Recognition

arXiv

引用

arXiv 2025年

作者： Zheng, Jinhui Liu, Zhiquan Si, Yain-Whar Li, Jianqing Zhang, Xinyuan Li, Xiaofan Huang, Haozhi Gong, Xueyuan School of Intelligent Systems Science and Engineering Jinan University Guangdong Province China College of Cyber Security Jinan University Guangdong Province China Department of Computer and Information Science Faculty of Science and Technology University of Macau China School of Computer Science and Engineering Macau University of Science and Technology China

Handwritten Paragraph Text Recognition (HPTR) is a challenging task in computer Vision, requiring the transformation of a paragraph text image, rich in handwritten text, into text encoding sequences. One of the most advanced models for this task is Vertical Attention Network (VAN), which utilizes a Vertical Attention Module (VAM) to implicitly segment paragraph text images into text lines, thereby reducing the difficulty of the recognition task. However, from a network structure perspective, VAM is a single-branch module, which is less effective in learning compared to multi-branch modules. In this paper, we propose a new module, named Re-parameterizing Vertical Attention Fusion Module (RVAFM), which incorporates structural re-parameterization techniques. RVAFM decouples the structure of the module during training and inference stages. During training, it uses a multi-branch structure for more effective learning, and during inference, it uses a single-branch structure for faster processing. The features learned by the multi-branch structure are fused into the single-branch structure through a special fusion method named Re-parameterization Fusion (RF) without any loss of information. As a result, we achieve a Character Error Rate (CER) of 4.44% and a Word Error Rate (WER) of 14.37% on the IAM paragraph-level test set. Additionally, the inference speed is slightly faster than VAN. © 2025, CC BY.

关键词： Parameterization

来源：评论

学校读者我要写书评

暂无评论

Integrating Multimodal Perception into Ground Mobile Robots

Integrating Multimodal Perception into Ground Mobile Robots

引用

IEEE International Conference on Autonomous Robot systems and Competitions (ICARSC)

作者： Ricardo B. Sousa Héber Miguel Sobreira João G. Martins Paulo G. Costa Manuel F. Silva António Paulo Moreira Faculty of Engineering University of Porto (FEUP) Porto Portugal INESC TEC - Institute for Systems and Computer Engineering Technology and Science Porto Portugal ISEP Polytechnic of Porto Porto Portugal

ISBN: (数字)9798331538606

ISBN: (纸本)9798331538613

Multimodal perception systems enhance the robustness and adaptability of autonomous mobile robots by integrating heterogeneous sensor modalities, improving long-term localisation and mapping in dynamic environments and human-robot interaction. Current mobile platforms often focus on specific sensor configurations and prioritise cost-effectiveness, possibly limiting the flexibility of the user to extend the original robots further. This paper presents a methodology to integrate multimodal perception into a ground mobile platform, incorporating wheel odometry, 2D laser scanners, 3D Light Detection and Ranging (LiDAR), and RGBD cameras. The methodology describes the electronics design to power devices, firmware, computation and networking architecture aspects, and mechanical mounting for the sensory system based on 3D printing, laser cutting, and bending metal sheet processes. Experiments demonstrate the usage of the revised platform in 2D and 3D localisation and mapping and pallet pocket estimation applications. All the documentation and designs are accessible in a public repository.

关键词： Laser radar Robot vision systems Wheels Three-dimensional printing Cameras Distance measurement Robustness Mobile robots Laser beam cutting Autonomous robots

来源：评论

学校读者我要写书评

暂无评论

Prediction of colorectal cancer based on taxonomic levels of microorganisms and discovery of taxonomic biomarkers using the Grouping-Scoring-Modeling (G-S-M) approach

引用

computers in Biology and Medicine 2025年 187卷 109813-109813页

作者： Bakir-Gungor, Burcu Temiz, Mustafa Canakcimaksutoglu, Beyza Yousef, Malik Department of Computer Engineering Faculty of Engineering Abdullah Gul University Kayseri38080 Turkey Department of Electrical and Computer Engineering Faculty of Engineering Abdullah Gul University Kayseri38080 Turkey Department of Bioengineering Faculty of Life and Natural Science Abdullah Gul University Kayseri38080 Turkey Department of Information Systems Zefat Academic College Zefat13206 Israel Zefat Academic College Israel

Colorectal cancer (CRC) is one of the most prevalent forms of cancer globally. The human gut microbiome plays an important role in the development of CRC and serves as a biomarker for early detection and treatment. This research effort focuses on the identification of potential taxonomic biomarkers of CRC using a grouping-based feature selection method. Additionally, this study investigates the effect of incorporating biological domain knowledge into the feature selection process while identifying CRC-associated microorganisms. Conventional feature selection techniques often fail to leverage existing biological knowledge during metagenomic data analysis. To address this gap, we propose taxonomy-based Grouping Scoring Modeling (G-S-M) method that integrates biological domain knowledge into feature grouping and selection. In this study, using metagenomic data related to CRC, classification is performed at three taxonomic levels (genus, family and order). The MetaPhlAn tool is employed to determine the relative abundance values of species in each sample. Comparative performance analyses involve six feature selection methods and four classification algorithms. When experimented on two CRC associated metagenomics datasets, the highest performance metric, yielding an AUC of 0.90, is observed at the genus taxonomic level. At this level, 7 out of top 10 groups (Parvimonas, Peptostreptococcus, Fusobacterium, Gemella, Streptococcus, Porphyromonas and Solobacterium) were commonly identified for both datasets. Moreover, the identified microorganisms at genus, family, and order levels are thoroughly discussed via refering to CRC-related metagenomic literature. This study not only contributes to our understanding of CRC development, but also highlights the applicability of taxonomy-based G-S-M method in tackling various diseases. © 2025 Elsevier Ltd

关键词： Colorectal cancer Human gut microbiome Machine learning Taxonomic biomarkers MicroBiomeGSM

来源：评论

学校读者我要写书评

暂无评论

Attackers Can Do Better: Over- and Understated Factors of Model Stealing Attacks

arXiv

引用

arXiv 2025年

作者： Oliynyk, Daryna Mayer, Rudolf Rauber, Andreas CDL AsTra Faculty of Computer Science University of Vienna Vienna Austria SBA Research Vienna Austria Institute of Information Systems Engineering Faculty of Informatics TU Wien Vienna Austria

Machine learning (ML) models were shown to be vulnerable to model stealing attacks, which lead to intellectual property infringement. Among other attack methods, substitute model training is an all-encompassing attack applicable to any machine learning model whose behaviour can be approximated from input-output queries. Whereas previous works mainly focused on improving the performance of substitute models by, e.g. developing a new substitute training method, there have been only limited ablation studies that try to understand the impact the strength of an attacker has on the substitute model’s performance. As a result, different authors came to diverse, sometimes contradicting, conclusions. In this work, we exhaustively examine the ambivalent influence of different factors resulting from varying the attacker’s capabilities and knowledge on a substitute training attack. Our findings suggest that some of the factors that have been considered important in the past are, in fact, not that influential;instead, we discover new correlations between attack conditions and success rate. In particular, we demonstrate that better-performing target models enable higher-fidelity attacks and explain the intuition behind this phenomenon. Further, we propose to shift the focus from the complexity of target models toward the complexity of their learning tasks. Therefore, for the substitute model, rather than aiming for a higher architecture complexity, we suggest focusing on getting data of higher complexity and an appropriate architecture. Finally, we demonstrate that even in the most limited data-free scenario, there is no need to overcompensate weak knowledge with unrealistic capabilities in the form of millions of queries. Our results often exceed or match the performance of previous attacks that assume a stronger attacker, suggesting that these stronger attacks are likely endangering a model owner’s intellectual property to a significantly higher degree than shown until now. Cop

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Video-Based Multiphysiological Disentanglement and Remote Robust Estimation for Respiration

引用

IEEE Transactions on Neural Networks and Learning systems 2025年第5期36卷 8360-8371页

作者： Shao, Hang Luo, Lei Qian, Jianjun Yan, Mengkai Gao, Shangbing Yang, Jian Nanjing University of Science and Technology PCA Laboratory Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education Jiangsu Key Laboratory of Image and Video Understanding for Social Security School of Computer Science and Engineering Nanjing210094 China Huaiyin Institute of Technology Faculty of Computer and Software Engineering Huaian223003 China

Remote noncontact respiratory rate estimation by facial visual information has great research significance, providing valuable priors for health monitoring, clinical diagnosis, and anti-fraud. However, existing studies suffer from disturbances in epidermal specular reflections induced by head movements and facial expressions. Furthermore, diffuse reflections of light in the skin-colored subcutaneous tissue caused by multiple time-varying physiological signals independent of breathing are entangled with the intention of the respiratory process, leading to confusion in current research. To address these issues, this article proposes a novel network for natural light video-based remote respiration estimation. Specifically, our model consists of a two-stage architecture that progressively implements vital measurements. The first stage adopts an encoder-decoder structure to recharacterize the facial motion frame differences of the input video based on the gradient binary state of the respiratory signal during inspiration and expiration. Then, the obtained generative mapping, which is disentangled from various time-varying interferences and is only linearly related to the respiratory state, is combined with the facial appearance in the second stage. To further improve the robustness of our algorithm, we design a targeted long-term temporal attention module and embed it between the two stages to enhance the network’s ability to model the breathing cycle that occupies ultra many frames and to mine hidden timing change clues. We train and validate the proposed network on a series of publicly available respiration estimation datasets, and the experimental results demonstrate its competitiveness against the state-of-the-art breathing and physiological prediction frameworks. © 2012 IEEE.

关键词： Physiological models

来源：评论

学校读者我要写书评

暂无评论

Detecting Advanced Persistent Threat Exfiltration with Ensemble Deep Learning Tree Models and Novel Detection Metrics

引用

IEEE Access 2025年 13卷 81803-81822页

作者： Cai, Xiaojuan Zhang, Haibo Ahmed, Chuadhry Mujeeb Koide, Hiroshi Kyushu University Department of Information Science and Technology Information Science and Electrical Engineering Fukuoka Japan Kyushu Institute of Technology Faculty of Computer Science and Systems Engineering Department of Artificial Intelligence Fukuoka Iizuka Japan Newcastle University Secure and Resilient Systems Group School of Computing Newcastle United Kingdom Kyushu University Section of Cyber Security for Information Systems Research Institute for Information Technology Fukuoka Japan

Advanced Persistent Threats (APTs) involve attackers maintaining a long-term presence on victim systems, leading to the stealthy exfiltration of sensitive data during network transfers. Despite existing methods to detect and halt APT data exfiltration, these attacks continue to pose significant threats to sensitive information and result in substantial commercial losses. Current approaches primarily focus on preemptive measures, which are insufficient once early-stage detection fails due to a lack of continuous monitoring. We propose an effective and efficient network monitoring method to address this gap and detect APT exfiltration during data transfer. Our approach assumes the presence of an undetected APT attacker within the victim system. We examine data exfiltration across three exfiltration traffic environments: exfiltration over command control channels, exfiltration over transfer size limitations, and their combinations. We introduce two detection metrics: Package Transfer Rate and Byte Transfer Rate. Utilizing these metrics, we measure network traffic, categorize APT attack environments, and train deep neural network models, named EDXGB, using ensembled decision trees to predict APT exfiltration. Our method is validated on two public datasets and compared against six baseline methods. Additionally, we simulate real-world exfiltration scenarios by creating three exfiltration traffic environments for each dataset. The results demonstrate that our method effectively detects APT exfiltration across various network environments, enhancing data protection and secure transfer. © 2013 IEEE.

关键词： Advanced Persistent Threat Data Exfiltration Deep Learning Privacy Preserving

来源：评论

学校读者我要写书评

暂无评论

Automated explainable and interpretable framework for anomaly detection and human activity recognition in smart homes

引用

Neural Computing and Applications 2025年 1-14页

作者： Kim, Tae Hoon Ojo, Stepehn Krichen, Moez Alamro, Meznah A. Mihoub, Alaeddine Sampedro, Gabriel Avelino School of Information and Electronic Engineering Zhejiang University of Science and Technology No. 318 Zhejiang Hangzhou China Department of Electrical and Computer Engineering College of Engineering Anderson Anderson University AndersonSC29621 United States Al-Baha University Al-Baha65528 Saudi Arabia ReDCAD Laboratory University of Sfax Sfax3038 Tunisia Department of Information Technology College of Computer and Information Science Princess Nourah Bint Abdul Rahman University Riyadh11564 Saudi Arabia Department of Management Information Systems College of Business and Economics Qassim University P.O. Box: 6640 Buraidah51452 Saudi Arabia College of Computer Studies De La Salle University 2401 Taft Ave. Malate Manila1004 Philippines Faculty of Information and Communication Studies University of the Philippines Open University Los Baños4031 Philippines

This paper presents a comprehensive framework for activity recognition and anomaly detection in smart home environments, targeting applications in convenience, efficiency, responsiveness, and healthcare. The proposed framework incorporates explainable artificial intelligence (XAI) to interpret feature impacts on learning models and optimize feature selection for activity recognition and anomaly detection. The study examines the recognition of two activity sets: Activities Set I, comprising 8 activities, and Activities Set II, comprising 11 activities. Various machine learning models (random forest (RF), decision tree (DT), and XGBoost (XGB)) are employed, along with an ensemble voting classifier (EVC) with these machine learning. EVC achieved 91% accuracy for Activities Set I and 90–91% accuracy for Activities Set II. The model demonstrated strong precision, recall, and F1-score across most classes. For anomaly detection, the Isolation Forest and H2O Isolation Forest Estimator algorithms are utilized to uncover irregularities in daily routines. The Isolation Forest identified anomalies in activities such as chores and desk-related tasks, while the H2O Isolation Forest Estimator detected no anomalies in either activity set. This study underscores the potential of the proposed framework to enhance smart home functionality and highlights avenues for future research and system improvements. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2025.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

MSConv: Multiplicative and Subtractive Convolution for Face Recognition

arXiv

引用

arXiv 2025年

作者： Zhou, Si Si, Yain-Whar Yuan, Xiaochen Li, Xiaofan Liu, Xiaoxiang Zhang, Xinyuan Lin, Cong Gong, Xueyuan School of Intelligent Systems Science and Engineering Jinan University Guangdong China Department of Computer and Information Science University of Macau China Faculty of Applied Sciences Macau Polytechnic University China

In Neural Networks, there are various methods of feature fusion. Different strategies can significantly affect the effectiveness of feature representation, consequently influencing the model’s ability to extract representative and discriminative features. In the field of face recognition, traditional feature fusion methods include feature concatenation and feature addition. Recently, various attention mechanism-based fusion strategies have emerged. However, we found that these methods primarily focus on the important features in the image, referred to as salient features in this paper, while neglecting another equally important set of features for image recognition tasks, which we term differential features. This may cause the model to overlook critical local differences when dealing with complex facial samples. Therefore, in this paper, we propose an efficient convolution module called MSConv (Multiplicative and Subtractive Convolution), designed to balance the model’s learning of salient and differential features. Specifically, we employ multi-scale mixed convolution to capture both local and broader contextual information from face images, and then utilize Multiplication Operation (MO) and Subtraction Operation (SO) to extract salient and differential features, respectively. Experimental results demonstrate that by integrating both salient and differential features, MSConv outperforms models that only focus on salient features. Copyright © 2025, The Authors. All rights reserved.

关键词： Convolution

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：