检索结果-内蒙古大学图书馆

Halo reduction multi-exposure image fusion technique

Multimedia Tools and Applications 2025年第13期84卷 12347-12370页

作者： Sharif, Rizwan Amin, Benish Sukhia, Komal Nain Department of Electrical Engineering Institute of Space Technology Islamabad Pakistan Department of Computer Science Institute of Space Technology Islamabad Pakistan

Multi-exposure image fusion (MEF) involves combining images captured at different exposure levels to create a single, well-exposed fused image. MEF has a wide range of applications, including low light, low contrast, night photography, medical imaging, and remote sensing. However, MEF methods often face issues like artifacts, halos around edges, color inconsistencies, noise amplification, and difficulty in preserving fine details. Moreover, assessing the quality of fused images objectively is complex due to the subjective nature of human perception. Solving the challenges is essential to developing efficient MEF techniques that produce high-quality results across various scenarios. The proposed technique introduces an approach to handling halo artifacts and implementing MEF. The Dense Scale-Invariant Feature Transform (DSIFT) is used to capture vital information about image brightness, texture, and edges from source images. Three weight maps are computed from the local mean, signal strength, and the global gradient for initial weight estimation. The local mean represents the brightness of specific image areas, signal strength preserves essential details like textures and edges while reducing image noise, and global gradient helps identify regions with significant pixel value shifts across multiple exposure images. The weight maps are then combined using a weighted average and a Gaussian smoothing filter is applied to reduce inherent noise and discontinuities in the original weights. Subsequently, pyramid decomposition is performed to generate a fused image. The efficiency of the proposed method is extensively tested on challenging multi-exposure image sequences. The results of the proposed approach demonstrate its superiority in both subjective evaluation and objective metrics like the MEF-Structural Similarity Index (MEF-SSIM), Natural Image Quality Evaluator (NIQE) and Gradient based performance measure (QAB/f). © The Author(s), under exclusive licence to Springer

关键词： Image fusion

来源：评论

学校读者我要写书评

暂无评论

K-Gate Lock: Multi-Key Logic Locking Using Input Encoding Against Oracle-Guided Attacks 25

K-Gate Lock: Multi-Key Logic Locking Using Input Encoding Ag...

引用

30th Asia and South Pacific Design Automation Conference, ASP-DAC 2025

作者： Lopez, Kevin Rezaei, Amin Computer Engineering and Computer Science Department California State University Long Beach United States

ISBN: (纸本)9798400706356

Logic locking has emerged to prevent piracy and overproduction of integrated circuits ever since the split of the design house and manufacturing foundry was established. While there has been a lot of research using a single global key to lock the circuit, even the most sophisticated single-key locking methods have been shown to be vulnerable to powerful SAT-based oracle-guided attacks that can extract the correct key with the help of an activated chip bought off the market and the locked netlist leaked from the untrusted foundry. To address this challenge, we propose, implement, and evaluate a novel logic locking method called K-Gate Lock that encodes input patterns using multiple keys that are applied to one set of key inputs at different operational times. Our comprehensive experimental results confirm that using multiple keys will make the circuit secure against oracle-guided attacks and increase attacker efforts to an exponentially time-consuming brute force search. K-Gate Lock has reasonable power and performance overheads, making it a practical solution for real-world hardware intellectual property protection. © 2025 Institute of Electrical and Electronics Engineers Inc.. All rights reserved.

关键词： Locks (fasteners)

来源：评论

学校读者我要写书评

暂无评论

Performance Evaluation of EfficientNet Models in Grape Leaf Disease Classification 1

Performance Evaluation of EfficientNet Models in Grape Leaf ...

引用

1st International Conference on Intelligent Systems and Computational Networks, ICISCN 2025

作者： Subramanya, S.G. Parkavi, A. Nitte Meenakshi Institute of Technology Department of Computer Science and Engineering Bengaluru India Ramaiah Institute of Technology Department of Computer Science and Engineering Bengaluru India

ISBN: (数字)9798331529246

ISBN: (纸本)9798331529246

Grape farming is a globally significant agricultural practice, but grapevines frequently encounter viral, fungal, and bacterial infections that compromise crop quality and yield. Conventional disease detection methods are labor-intensive and require significant expertise. This study investigates deep learning techniques, focusing on EfficientNet models, to automate grape leaf disease identification and classification. The transfer learning approach was applied to train and fine-tune EfficientNet models on a publicly available dataset with four classes: black rot, leaf blight, esca, and healthy leaves. A model performance was measured through accuracy and loss metrics. EfficientNetB0 achieved accuracy of 92.45% and 96.73%, and loss of 0.197 and 0.098, on training and validation data, respectively. However, EfficientNetB5 and EfficientNetB7 gave validation accuracies of 96.07% and 96.39%, respectively. The accuracy and loss trends demonstrate that EfficientNetB0 outperforms the other two counterparts. This approach highlights the potential of EfficientNetB0 for accurate, efficient grape disease detection and classification. © 2025 IEEE.

关键词： Plant diseases

来源：评论

学校读者我要写书评

暂无评论

Control Structures and Algorithms for Force Feedback Bilateral Teleoperation Systems: A Comprehensive Review

引用

computer Modeling in engineering & sciences 2025年第2期142卷 973-1019页

作者： Jiawei Tian Yu Zhou Lirong Yin Salman A.AlQahtani Minyi Tang Siyu Lu Ruiyang Wang Wenfeng Zheng Department of Computer Science and Engineering Major in Bio Artificial IntelligenceHanyang UniversityAnsan-si15577Republic of Korea Department of Geography and Anthropology Louisiana State UniversityBaton RougeLA 70803USA Department of Computer Engineering College of Computer and Information SciencesKing Saud UniversityRiyadh11574Saudi Arabia School of Automation University of Electronic Science and Technology of ChinaChengdu610054China

Force feedback bilateral teleoperation represents a pivotal advancement in control technology,finding widespread application in hazardous material transportation,perilous environments,space and deep-sea exploration,and healthcare *** paper traces the evolutionary trajectory of force feedback bilateral teleoperation from its conceptual inception to its current *** elucidates the fundamental principles underpinning interaction forces and tactile exchanges,with a specific emphasis on the crucial role of tactile *** this review,a quantitative analysis of force feedback bilateral teleoperation development trends from 2011 to 2024 has been conducted,utilizing published journal article data as the primary source of *** review accentuates classical control frameworks and algorithms,while also delving into existing research advancements and prospec-tive breakthrough ***,it explores specific practical scenarios ranging from intricate surgeries to hazardous environment exploration,underscoring the technology’s potential to revolutionize industries by augmenting human manipulation of remote *** underscores the pivotal role of force feedback bilateral teleoperation as a transformative human-machine interface,capable of shaping flexible control strategies and addressing technological *** research endeavors in force feedback bilateral teleoperation are expected to prioritize the creation of more immersive experiences,overcoming technical hurdles,fortifying human-machine collaboration,and broadening application domains,particularly within the realms of medical intervention and hazardous *** the continuous progression of technology,the integration of human intelligence and robotic capabilities is expected to produce more innovations and breakthroughs in the field of automatic control.

关键词： Force feedback bilateral teleoperation robot development history application prospect

来源：评论

学校读者我要写书评

暂无评论

Classification of Brain-MRI Images Using a Stacked-Deep-Network Ensemble Model into Multiple Region-Based Classes 1st

Classification of Brain-MRI Images Using a Stacked-Deep-Netw...

引用

1st International Conference on Intelligent Systems, Blockchain, and Communication Technologies, ISBCom 2024

作者： Kumar, Deepika Srivastava, Varun Gupta, Shilpa Jamil, Akhtar Department of Computer Science and Engineering Bharati Vidyapeeth’s College of Engineering New Delhi India Department of Computer Science and Information Technology Jaypee Institute of Information Technology Uttar Pradesh Noida India Department of Computer Science and Engineering JIMS Engineering Management and Technical Campus Uttar Pradesh Greater Noida India Department of Computer Science National University of Computer and Emerging Sciences Islamabad Pakistan

ISBN: (纸本)9783031823763

Brain tumour besides being lethal can also affect other human organs on a long-term basis if not detected at an early stage. Based on the region of its presence and speed of growth, it can be classified as Glioma, Meningioma. or Pituitary. In the given work, a Stacked Neural Net approach for the classification of brain tumour is presented. The images are first pre-processed using Histogram equalization and One Hot Encoding. Thereby, the improved images are fed into 3 popular pre-trained deep networks viz. Res-Net-50, VGG-16, and Xception as level-0 models in parallel. The features extracted are then fed into a neural network as level 1 learner for the final classification. A publicly available Figshare dataset has been used for testing the performance. The suggested approach has been thoroughly evaluated against its base models using accuracy, precision, and recall metrics and also the performance has been analysed using accuracy and loss curves along with epochs. The suggested model achieved a validation and training accuracy of 97.3% and 98.12% respectively. It has been observed that the suggested model referred here as Stacked-Neural-Net outperforms when compared to the conventional deep learning methods and state-of-the-art methods. An improvement of 3.35%, 1.3% and 1.7% is achieved over Inception ResNetV1, a basic CNN and Alex-Net CNN models. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Fall Detection and Alerts in Hearing Aids for Deaf using Generative AI 2

Fall Detection and Alerts in Hearing Aids for Deaf using Gen...

引用

2nd International Conference on Computational Intelligence, Communication Technology and Networking, CICTN 2025

作者： Tripathi, Bharat Lal Saini, Madan Jain, Manish Saxena, Apoorva Pal, Narendra Department of Computer Application Allenhouse Institute of Technology Kanpur India Department of Computer Science and Engineering Apex Institute of Technology Chandigarh University Punjab India Department of Computer Application Allenhouse Business School Kanpur India Department of Computer Science & Engineering Axis Institute of Technology & Management Kanpur India Department of Computer Science & Engineering Invertis University Bareilly India

ISBN: (纸本)9798331530389

Individuals with sensorineural hearing loss often experience difficulty comprehending speech when background noise is present. This paper investigates the extent of this problem in various listening scenarios and with different types of background sounds. Additionally, the section explores the contributing factors to this challenge such as reduced audibility, reduced frequency selectivity, loudness recruitment, and dead regions in the cochlea where there are no surviving inner hair cells and or neurons. The sections evaluate different techniques for compensating for the effects of these factors. While signal-processing methods using the output of a single microphone to compensate for reduced frequency selectivity have had limited success, techniques utilizing multiple microphones have proven effective. Amplitude compression has also been effective in compensating for the effects of loudness recruitment, enabling individuals to understand speech across a wide range of sound levels. Next, the paper delves into the technical aspects of implementing fall detection systems in hearing aids, learning algorithms are also investigated for their potential in enhancing fall detection accuracy and minimizing false alarms. Furthermore, the importance of user-friendly interfaces and customization features in ensuring the effectiveness and acceptance of fall detection alerts is also highlighted. The paper also addresses the ethical and privacy considerations associated with fall detection in hearing aids. It emphasizes the need for transparent data handling practices and user consent, while also addressing potential concerns related to user autonomy and intrusion. © 2025 IEEE.

关键词： Data privacy

来源：评论

学校读者我要写书评

暂无评论

Optimized COVID-19 detection using sparse deep learning models from multimodal imaging data

引用

Multimedia Tools and Applications 2025年第13期84卷 11451-11474页

作者： Moradi, MohammadMahdi Hassanzadeh, Alireza Haghanifar, Arman Ko, Seok Bum Faculty of Computer Science and Software Engineering Concordia University MontrealQC Canada Faculty of Electrical Engineering Shahid Beheshti University Tehran Tehran Iran Department of Electrical & Computer Engineering University of Saskatchewan SaskatoonSK Canada

This paper explores the global spread of the COVID-19 virus since 2019, impacting 219 countries worldwide. Despite the absence of a definitive cure, the utilization of artificial intelligence (AI) methods for disease diagnosis has demonstrated commendable effectiveness in promptly diagnosing patients and curbing infection transmission. The study introduces a deep learning-based model tailored for COVID-19 detection, leveraging three prevalent medical imaging modalities: computed tomography (CT), chest X-ray (CXR), and Ultrasound. Various deep Transfer Learning Convolutional Neural Network-based (CNN) models have undergone assessment for each imaging modality. For each imaging modality, this study has selected the two most accurate models based on evaluation metrics such as accuracy and loss. Additionally, efforts have been made to prune unnecessary weights from these models to obtain more efficient and sparse models. By fusing these pruned models, enhanced performance has been achieved. The models have undergone rigorous training and testing using publicly available real-world medical datasets, focusing on classifying these datasets into three distinct categories: Normal, COVID-19 Pneumonia, and non-COVID-19 Pneumonia. The primary objective is to develop an optimized and swift model through strategies like Transfer Learning, Ensemble Learning, and reducing network complexity, making it easier for storage and transfer. The results of the trained network on test data exhibit promising outcomes. The accuracy of these models on the CT scan, X-ray, and ultrasound datasets stands at 99.4%, 98.9%, and 99.3%, respectively. Moreover, these models’ sizes have been substantially reduced and optimized by 51.93%, 38.00%, and 69.07%, respectively. This study proposes a computer-aided-coronavirus-detection system based on three standard medical imaging techniques. The intention is to assist radiologists in accurately and swiftly diagnosing the disease, especially during the screen

关键词： COVID-19

来源：评论

学校读者我要写书评

暂无评论

Sem-AI: A Unique Framework for Sentiment Analysis and Opinion Mining Using Social Network Data

引用

SN computer science 2025年第2期6卷 1-23页

作者： Maruthupandi, J. Sivakumar, S. Kumar, V. Senthil Srikaanth, P. Balaji Department of Computer Science and Engineering New Horizon College of Engineering Karnataka Bangalore India Department of Computer Science and Engineering SRM Institute of Science and Technology Tamilnadu Trichy India Department of Networks and Communication SRM Institute of Science and Technology Tamilnadu Chennai India Department of Computer Science and Engineering Vel Tech Rangarajan Dr. Sagunthala R & D Institute of Science and Technology Chennai India

Social media has become an essential forum for people to share their thoughts and sentiments owing to the quick rise in mobile technology. Business and political organizations might benefit from understanding public sentiment when making strategic decisions. Challenges in the area include handling noisy and incomplete data, variation in language use across domains, and integration into existing systems of models developed for analysis. In these ways, they can seriously hamper the accuracy and generalizability of sentiment analysis models. In view of the challenges, this work has developed a new and innovative framework of AI-based semantic analysis for assessing and classifying user views from social data. In spite of this, sentiment analysis is crucial for determining the opposing perspectives of the general public's sentiments. This work constructed a new and innovative framework called Semantic analysis based on AI (SemAI) for assessing and categorizing user views from social data. With high efficiency, a collection of intelligence algorithms in the stages of feature extraction, optimization, and sentiment prediction are applied here. The Degree of Correlation Network Model (DCNM) is used in this work to extract the subset of dependable and most-required features from social network data. The most recent meta-heuristics model, Heap based Optimization (HbO), is used to reduce the dimensionality of retrieved features. Furthermore, the unique classification algorithm, Self-Attention based Deep Analyzing Network (SA-DAN), is used in this work to precisely identify and categorize attitudes into positive, neutral, and negative classifications. This present study used a variety of most recent social corpus data to investigate and evaluate the Sem-AI model's outputs. The SemAI model achieved impressive results, with accuracy, precision, and F1-scores of 99%, 98.9%, and 99%, respectively, across various social corpus datasets. © The Author(s), under exclusive licence to S

关键词： Classification Degree of correlation network model (DCNM) Heap based optimization (HbO) Opinion mining Self-attention based deep analyzing network (SA-DAN) Semantic analysis Social data

来源：评论

学校读者我要写书评

暂无评论

Enhancing reliability in coronary artery disease prediction using ensemble learning for key attribute identification

引用

Life Cycle Reliability and Safety engineering 2025年 1-14页

作者： Upadhyay, Sachin Sagar, Anil Kumar Roy, Nihar Ranjan Department of Computer Science and Engineering School of Engineering and Technology Sharda University Knowledge Park III Greater Noida India Department of Computer Science and Engineering School of Engineering and Technology VIPS-TC Delhi Pitampura India

Cardiovascular disease (CAD) is a significant public health concern, affecting a large population worldwide. Early diagnosis and management of CAD can minimize the risk of acute myocardial infarction and improve patient outcomes. Assessment tools like SBP, cholesterol, pulse rate, and ST segment depression can help identify causes early and manage them effectively. Management includes medication therapy, healthy dietary habits, and exercise. Several machine learning (ML) methodologies have been researched to enhance CAD predictions, including AdaBoost, ANNs, J48, Decision Tree, K Nearest Neighbor (KNN), Naïve Bayes, and Random Forest. However, single models still lack sufficient capacity to address the complexity and flexibility of CAD. Ensemble learning, which uses multiple classifiers to boost predictability, has been employed to address these issues. The system was developed after benchmarking it with multiple classifiers on a Cleveland cardiac disease dataset. The ensemble method showed a 92.11% accuracy rate, far better than the highest performing classifier operating individually. This suggests the possibility of practical solutions for CAD prediction using ensemble methods, enabling precise early diagnosis and efficient targeted treatment. Comparing ensemble learning for CAD predictors reveals how these approaches can revolutionize medicine by enabling early diagnosis and personalized treatment plans. There is a need to further develop these methods for clinical application, such as creating practical tools for easier application by healthcare workers and integrating sophisticated techniques. In conclusion, ensemble learning methods represent significant advancements in CAD prediction, with superior performance in identifying critical attributes and enhancing predictive accuracy. As healthcare evolves with the integration of intelligent technologies, the adoption of ensemble learning methods holds great promise for enhancing patient outcomes and reducing the

关键词： ANN CAD Decision Trees Ensemble KNN Machine learning Random Forest

来源：评论

学校读者我要写书评

暂无评论

CUET_HateShield@NLU of Devanagari Script Languages 2025: Transformer-Based Hate Speech Detection in Devanagari Script Languages 1

CUET_HateShield@NLU of Devanagari Script Languages 2025: Tra...

引用

1st Workshop on Challenges in Processing South Asian Languages, CHiPSAL 2025, co-located with the 31st International Conference on Computational Linguistics, COLING 2025

作者： Aodhora, Sumaiya Rahman Ahsan, Shawly Hoque, Mohammed Moshiul Department of Computer Science and Engineering Chittagong University of Engineering and Technology Bangladesh

ISBN: (纸本)9798891762015

Social networks have become essential platforms for information exchange and free expression. However, their open nature also facilitates the spread of harmful content, such as hate speech, cyberbullying, and offensive language, which pose significant risks to social well-being. This study focuses on developing an automated system to detect hate speech in Devanagari script languages, enabling efficient moderation and timely intervention. Our approach leverages a fine-tuned transformer model for classifying offensive content. We experimented with various machine learning (ML) techniques, including Logistic Regression (LR), Support Vector Machines (SVM), and Random Forest (RF), as well as deep learning (DL) architectures such as CNN, BiLSTM, and CNN-BiLSTM. Additionally, we evaluated transformer-based models, including IndicBERT, m-BERT, MuRIL, Indic-SBERT, and XLM-R. Among these, the fine-tuned XLM-R model delivered the best performance, achieving a macro f1-score of 0.74, demonstrating its effectiveness in detecting hate speech in Devanagari script languages. However, the model submitted for the shared task achieved a macro f1-score of 0.73, ranking 13th in the subtask. © 2025 International Committee on Computational Linguistics.

关键词： Decision trees

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：