检索结果-内蒙古大学图书馆

PRF: deep neural network compression by systematic pruning of redundant filters

Neural Computing and Applications 2024年第33期36卷 20607-20616页

作者： Sarvani, C.H. Ghorai, Mrinmoy Basha, S. H. Shabbeer Computer Vision Group Indian Institute of Information Technology Andhra Pradesh Sri City Chittoor517646 India School of Computer Science and Engineering RV University Karnataka Bengaluru560059 India

In deep neural networks, the filters of convolutional layers play an important role in extracting the features from the input. Redundant filters often extract similar features, leading to increased computational overhead and larger model size. To address this issue, a two-step approach is proposed in this paper. First, the clusters of redundant filters are identified based on the cosine distance between them using hierarchical agglomerative clustering (HAC). Next, instead of pruning all the redundant filters from every cluster in single-shot, we propose to prune the filters in a systematic manner. To prune the filters, the cluster importance among all clusters and filter importance within each cluster are identified using the 1-norm based criterion. Then, based on the pruning ratio filters from the least important cluster to the most important ones are pruned systematically. The proposed method showed better results compared to other clustering-based works. The benchmark datasets CIFAR-10 and ImageNet are used in the experiments. After pruning 83.92% parameters from VGG-16 architecture, an improvement over the baseline is observed. After pruning 54.59% and 49.33% of the FLOPs from ResNet-56 and ResNet-110, respectively, both showed an improvement in accuracy. After pruning 52.97% of the FLOPs, the top-5 accuracy of ResNet-50 drops by only 0.56 over ImageNet. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Test Suite Minimization for Regression Testing Using Keyword Match Technique 7th

Test Suite Minimization for Regression Testing Using Keyword...

引用

7th International Conference on Information and Communication Technology for Competitive Strategies , ICTCS 2022

作者： Freeda, R. Adline Rajendran, P. Selvi Department of Computer Science and Engineering Hindustan Institute of Science and Technology Chennai India

ISBN: (纸本)9789811996375

Regression testing is a crucial phase of the testing practice as it ensures stability and functionality of the system. It repeats test cases from the previous versions to ensure that the original function of the system is not affected by the new changes. The agile methodologies had gained popularity in the recent years which stresses the key importance of regression testing in protecting software quality, and regression testing has made significant development in the recent years. In an agile framework, the ideal scenario for regression testing is to run a regression suite of test cases after each sprint and release, which consumes money and time. In the proposed system, using the keyword match technique, we provide a test case coverage mechanism that predicts the maximum coverage of test cases over the requirements in this work. Following that, test cases are chosen based on maximum coverage. The approach improves both prioritization and decision-making. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

Vision Based Detection and Analysis of Human Activities 7

Vision Based Detection and Analysis of Human Activities

引用

7th International Conference on Trends in Electronics and Informatics, ICOEI 2023

作者： Ravipati, Abhiram Kondamuri, Rakesh Krishna Mary Posonia, A. Albert Mayan, J. Sathyabama Institute of Science and Technology Department of Computer Science and Engineering Chennai India

ISBN: (纸本)9798350397284

These days, it's not uncommon to see video cameras installed to keep an eye on pedestrians and motorists alike in a variety of public spaces. The proliferation of camera footage necessitates the creation of some means of deducing the nature of the activity captured on film. Due to the widespread availability of acquisition devices like handsets & camcorders, HAR may be used in a wide variety of contexts. The proliferation of electronic gadgets and software has been matched by a revolution in data extraction made possible by breakthroughs in AI. Difficulties such backdrop congestion, partly blockage, variations in size, perspective, illumination, & look, make it difficult to recognize human actions in video frame or still photographs. Multimodal recognition is necessary in several fields, such as CCTV, human computer interaction for characterizing human behavior. This study surveys the most current and cutting-edge findings from the study of how to categorize human actions. This research proposes a taxonomy for studying human activity and explore the benefits and drawbacks of this diverse set of approaches. Disabled people's daily routines, including resting, moving, traveling down & up the stairs, speaking, & laying, have been frequently tracked using cellphones. Human motion detection often makes use of well-established ML & DL techniques include CNN, & LSTM Network. © 2023 IEEE.

关键词： Video cameras

来源：评论

学校读者我要写书评

暂无评论

Photovoltaic Panel Quality Evaluation Based on Power Generated Involving Copper Indium Gallium Selenide Solar Cells

Photovoltaic Panel Quality Evaluation Based on Power Generat...

引用

2024 International Conference on Sustainable Computing and Integrated Communication in Changing Landscape of AI, ICSCAI 2024

作者： Srivastava, Yashi Almusawi, Muntather Badhoutiya, Arti Dhasmana, Gulshan Jain, Alok Aishwarya, Bk Boob, Nandini Shirish Srivastava, Arun Pratap Lloyd Institute of Engineering & Technology Greater Noida India Department of Computers Techniques Engineering College of Technical Engineering The Islamic University Najaf Iraq Department of Electrical Engineering GLA University Mathura India Department of Computer Science & Engineering Graphic Era Deemed to be University Dehradun India Lovely Professional University Phagwara India Institute of Aeronautical Engineering Dundigal Hyderabad India Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Tamilnadu Chennai India

ISBN: (纸本)9798350355338

When it comes to energy-efficient manufacturing, this report tackles some of the most important concerns raised by companies and academic institutions. Both production management and the integration of information and communication technology systems with respect to energy efficiency performance criteria are encompassed in the research question (RQ). The requirements of the corporate sector were considered alongside solutions that had previously been detailed in the literature. Findings show that methods exist for keeping tabs on, managing, and improving production processes. Organizational, facility, and procedure-level energy management in manufacturing is an area where these methods are seldom effective. A gap persists in industrial organizations between possible solutions and their actual execution. To close the gap between theory and practise, research into effective and efficient energy management in manufacturing is necessary. Managers in the manufacturing sector may use this data to make more informed, faster business decisions based on real energy usage. Key performance indicators (KPIs) for energy efficiency should be defined and specified in a way that allows for quantitative measurement at the plant level. Seeing how energy efficiency stacks up against other metrics is essential for decision-makers looking at production performance areas from multiple angles. © 2024 IEEE.

关键词： Quality control

来源：评论

学校读者我要写书评

暂无评论

An Optimized Hybrid Soil Texture Classification Model using Stacked Sparse Autoencoder 1

An Optimized Hybrid Soil Texture Classification Model using ...

引用

1st IEEE International Conference on Emerging Research in Computational science, ICERCS 2023

作者： Prabavathi, R. Chelliah, Balika J Srm Institute of Science and Technology Department of Computer Science and Engineering Chennai India

ISBN: (纸本)9798350359756

With effective soil management, which lowers the costs associated with soil erosion, improves water utilisation, and other factors, soil data analysis can offer vital information to raise agricultural yield and nutrient utilisation efficiency. An unsupervised method of data analysis called clustering finds similar groups of objects by analysing attribute values. The dynamic subject of soil classification starts with the model and moves on to explain the classes and, at the end, apply the knowledge gained from the field. In light of this, the study presents a novel clustering and classification-based hybrid soil texture classification model (HSTC). Soil data will be clustered and categorised into multiple classes by the proposed (HSTC) model. Hydrological soil groupings (HSG) of various kinds are used as classes to accomplish this. Based on characteristics like the proportion of clay, silt, and sand, the HGS can be categorised. The density-based clustering (DBSCAN) algorithm is used to group the soil texture data. The stacked sparse auto encoder (SSAE) model is used to categorise the clustered data. In addition, the parameter values of the SSAE technique are determined using a new barnacle mating optimizer (BMO). The HSTC model has achieved the best results, with accuracy of 95.66%. The experimental validation of the proposed HSTC model on soil texture dataset revealed that the suggested model outperformed previous techniques. © 2023 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Implementation of smart sensors with Swarm Intelligence for Disease Diagnosis in Smart Environments 1

Implementation of smart sensors with Swarm Intelligence for ...

引用

1st IEEE International Conference on Emerging Research in Computational science, ICERCS 2023

作者： Abirami, G. Poornima, D. Premalatha, J. Sarojini Ridhi Mayura Shivani, R. Shree, T. Bhanu Sathyabama Institute of Science & Technology Department of Computer Science and Engineering Chennai India

ISBN: (纸本)9798350359756

Disease diagnosis in smart settings has been transformed by the fast development of technology, especially the Internet of Things (IoT) and Swarm Intelligence. Various health-related data may be continually monitored and collected in smart environments due to the abundance of sensors. However, this data's vast quantity and complexity complicates accurate illness diagnosis and early management. Inspired by the coordinated actions of flocks of birds and swarms of insects, Swarm Intelligence provides a novel paradigm for analyzing and understanding this mountain of data. Hence, this paper proposes a Swarm Intelligence-enabled Disease Diagnosis (SI-DD) in smart environments using IoT-based sensors. Smart systems use algorithms motivated by swarm behaviour to assess data from several sensors in concert, allowing for faster, more accurate illness diagnosis. As a shape of collective swarm intelligence, Ant Colony Optimization (ACO) enables make healthcare treatments more effective. As a result, sufferers get care tailor-made to their unique desires, thinking about their scientific history and the kingdom of the healthcare machine. This article examines the most distinguished makes use of the Internet of Things and Swarm Intelligence in healthcare. It highlights how those techniques have already shown their promise for higher contamination detection. With this in thoughts, it's clear that the usage of clever sensors prepared with Swarm Intelligence for infection detection in clever settings is a viable path toward enhancing healthcare results and the first-rate of existence for sufferers. This novel approach has the capability to revolutionize contamination surveillance, prognosis, and control, leading to greater cost-effective scientific services and higher results for sufferers. © 2023 IEEE.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Optimizing Plant Disease Detection Using Convolutional Neural Networks and Docker Environment

Optimizing Plant Disease Detection Using Convolutional Neura...

引用

2025 International Conference on Computing for Sustainability and Intelligent Future, COMP-SIF 2025

作者： Rachana, C.V. Jai Arul Jose, G. Bms Institute of Technology and Management Department of Computer Science and Engineering Bengaluru India

ISBN: (数字)9798331538538

ISBN: (纸本)9798331538538

In numerous nations, agriculture serves as the primary and most significant source of income generated domestically. Plant illnesses caused by bacteria, viruses, and fungi are among the many, and they can be quite expensive for agricultural companies worldwide. Crop security both in terms of quantity and quality is crucial for monitoring plant disease. Therefore, it is essential to recognize plant diseases. There are certain plant parts where the disease state is clearly visible. However, usually, the infection is seen in specific plant leaves. Many researchers use computing techniques to automatically detect plant illness using photos of their leaves. Farmers might also profit fromthese strategies by obtaining quick and suitable measures to do in both agricultural yield and quality. Furthermore, some molecular methods have been developed to reduce and even eliminate the harm posed by pathogens. Thus, this evaluationaids the researcher to identify automatically the plant illness. Also enhancing the learning and provide specific methods for disease identification in order to prevent illness. © 2025 IEEE.

关键词： Crops

来源：评论

学校读者我要写书评

暂无评论

End to End Model to Reduce the Inference, Jamming, and to Increase the Trust from the Compromised Secondary Nodes in Cognitive Radio Networks 2

End to End Model to Reduce the Inference, Jamming, and to In...

引用

2nd IEEE International Conference on Intelligent and Innovative Technologies in Computing, Electrical and Electronics, ICIITCEE 2024

作者： Dhananjaya, Shashank Narender, M. Sunitha, R. The National Institute of Engineering Department of Information Science and Engineering Mysore India The National Institute of Engineering Department of Computer Science and Engineering Mysore India B. N. M. Institute of Technology Dept. of Artificial Intelligence and Machine Learning India

ISBN: (纸本)9798350306415

Wireless technology has advanced dramatically in recent years. Multimedia applications have exploded in popularity over wireless networks. The demand for radio spectrum has expanded as the demand for large capacity for multimedia applications has grown. The traditional fixed spectrum assignment approach has been challenged by the shortage of radio spectrum. As a result, Cognitive Radio (CR) has arisen as a new unique option for addressing spectrum underutilization by allowing users to access unused spectrum bands on the fly. In cognitive radio networks, radio spectrum scarcity, spectrum access, signal interference, security at various layers, trust, jamming, and spectrum management have all become key issues. Proposing a unique approach to address the interference problem produced by secondary nodes that have been hacked. To fulfil the need, we develop a mathematical modelling of the system that incorporates intelligence in the SU to cope with intelligent jammer. Intelligent jammer is cognitive in nature, having the ability to learn, to reason and to adjust its strategies against SU for maximum damage to the CRN. Many researchers have been working on reducing inference and the effect of jamming in the network. But still it is not resulted with the proper solution for the jamming problem in cognitive radio wireless sensor network. Hence proposing the methodology aiming to minimize the effect of Jamming in Cognitive Radio Wireless Sensor-Networks. At the end, the proposed research is aiming to solve the security problem to increase the trust in the cognitive radio network. The secondary users must be able to detect and forecast spectrum holes in the CR environment. The overall result shows the improved network performance by reducing the inference, jamming, spectrum sensing and trust issues in cognitive radio network. © 2024 IEEE.

关键词： Cognitive Radio Network Inference Jamming Q-Learning Software Defined Radio Trust

来源：评论

学校读者我要写书评

暂无评论

Improving Face Recognition for Smart Home Security: A Study of Resnet, Siamese Networks, and Attention Mechanisms

Improving Face Recognition for Smart Home Security: A Study ...

引用

2025 International Conference on Computing for Sustainability and Intelligent Future, COMP-SIF 2025

作者： Sood, Hemaang Nikhil, B.V.S. Dengla, Palak Saeed, Sanya Nagabhushan, S.V. Bms Institute of Technology and Management Department of Computer Science and Engineering Bengaluru India

ISBN: (纸本)9798331538538

This study presents a smart doorbell system designed to improve home security through advanced deep learning techniques. At its core, the system utilizes the YOLOv5 (You Only Look Once) algorithm for real-time face detection, ensuring accurate and efficient visitor identification. To further strengthen security, the system employs multifactor authentication (MFA), integrating face recognition alongside speech recognition and or/a manual keypad. The study evaluates three neural network architectures for face recognition-ResNet with cosine similarity, a Siamese network with CNN, and an attention-enhanced Siamese network-to determine the most effective model for secure and reliable authentication. Performance comparisons highlight progressive improvements in accuracy and robustness,. Additionally, a convenient interface enables seamless remote monitoring and control, enhancing both security and convenience. By addressing key limitations of existing smart security systems, this research provides insight into improving intelligent home access control. © 2025 IEEE.

关键词： Authentication

来源：评论

学校读者我要写书评

暂无评论

Competent Diagnosis in Agriculture: Federated Learning CNNs for Avocado Leaf Diseases

Competent Diagnosis in Agriculture: Federated Learning CNNs ...

引用

2024 Asia Pacific Conference on Innovation in Technology, APCIT 2024

作者： Suryavanshi, Ankita Mehta, Shiva Kumar, Mukesh Thapliyal, Siddhant Jain, Vishal Chitkara University Institute of Engineering and Technology Chitkara University Punjab India Computer Science and Engineering Graphic Era Hill University Uttarakhand Dehradun India Computer Science and Engineering Graphic Era Deemed to Be University Uttarakhand Dehradun India Computer Science and Engineering Sharda School of Engineering and Technology Sharda University U.P. Greater Noida India

ISBN: (纸本)9798350361537

This research paper explores the novel application of federated learning with Convolutional Neural Networks (CNN) for avocado leaf disease classification across the four severity categories. Based on the details from four unique customers, the analysis shows how federated learning can capitalize in agricultural environments around local data variations and improve the global model's accuracy. Our approach is the only one to tackle the classification issues with four disease severity classes. However, the fusion of federated learning with CNNs secures data privacy by processing locality at each client and promotes model generalisability over the various datasets. In the results, for each client (kg-1 to kg-4), there are signs of a good level of precision, recall, F1-score support, and accuracy in the disease classification. In particular, the client kg-1 determined a macro average of 96.73%, a weighted average of 96.80%, and a micro average of 96.80% with an accuracy equal to. Client kg-3 was optimized with a macro-average efficiency score of 98.95%, a weighted average for performance scoring set to 0.7, and a micro-average. The peak performance was finally reached for client kg4 with a macro average of 99.15%, a weighted average of 98.73%, and a micro average age of ninety-six while maintaining accuracy. These results show the high performance of the federated learning CNN model in diagnosing avocado leaf diseases with great precision and present many opportunities to use this methodology for other agricultural applications. © 2024 IEEE.

关键词： Federated learning

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：