检索结果-内蒙古大学图书馆

The use of UAV for the Identification of Flooded and Non-Flooded Region in Post Flood Scenarios using Deep Learning Techniques

引用

Procedia computer science 2025年 260卷 533-539页

作者： Rajesh Kumar Sinha Mohit Kushwaha Jaytrilok Choudhary Dhirendra Pratap Singh Manish Pandey Department of Computer Science and Engineering MANIT Bhopal Madhya Pradesh India (462003)

Floods are a form of unforeseen disaster which may seriously compromise human health and safety and also damage infrastructure, property, and ecosystems. Post flood management is process which involves better flood image segmentation, the mapping process, assistance with relief, and evacuation these can be enhanced by the adoption of technology. In computer vision and image processing, the segmentation of image is crucial task, it is a method which determine different types of objects in an image through defining labels to each and every pixel in a target category according to its semantics. Use of Unmanned Aerial Vehicle (UAV) is more reliable way to capturing images and video of the flood regions. This study presents a comparative analysis of the achievement of DeepLabv3+ in segmentation flooded and non-flooded regions in post-flood UAV images. The models are assessed based on key performance metrics: Accuracy, Precision, Recall, Intersection over Union (IoU) and Dice Loss. Among these, DeepLabv3+ emerged as the most effective, achieving the highest accuracy (0.9494), precision (0.9565) and IoU (0.7686), indicating its superior ability to correctly segment flood images and minimize false positives. Although its recall (0.8781) was slightly lower than that of DeepLabv3, it still demonstrated robust flood area identification. The lower Dice Loss (0.1332) further corroborates its superior segmentation quality with better segmentation.

关键词： Post Flood Image UAV Images image segmentation DeepLabv3+

来源：评论

学校读者我要写书评

暂无评论

High-Performance Image Restoration with SpectraNet: A Lightweight CNN Architecture

引用

Procedia computer science 2025年 258卷 2268-2277页

作者： Nitesh Patnaik Rishi Raj Piyush Kumar Singh Shreekara Mahapatra Debashis Pani Ravi Shankar Singh Dept. of Computer Science and Engineering Kalinga Institute Of Industrial Technology Bhubaneswar 751024 India Dept. of Computer Science Central University of South Bihar Bihar 824236 India Dept. of Computer Science and Engineering Indian Institute of Technology BHU Varanasi 221005 India

Deep learning-based image restoration faces significant challenges when deployed on resource-constrained platforms due to the computational demands and large number of parameters in existing models. The high computational load and extensive memory requirements make it difficult to implement these models on devices with limited processing power and storage capacity, such as mobile phones and embedded systems. Additionally, maintaining real-time performance while ensuring high-quality image restoration is a critical challenge, as traditional deep learning models often fail to meet the stringent latency and efficiency requirements of such platforms. This paper introduces a new, lightweight convolutional neural network (CNN) architecture tailored for efficient pixelation detection and restoration. Our approach combines a pre-trained MobileNetV2 with finetuning for pixelation detection, and SpectraNet, a architecture incorporating depth-wise separable convolutions for image restoration. The proposed architecture was trained on the HQ-50k dataset, with 70% of the data used for training and 30% for testing. It achieved a Peak Signal-to-Noise Ratio (PSNR) of 17-23, outperforming 10 state-of-the-art architectures in both efficiency and image quality. This design addresses the dual problems of high computational load and extensive memory usage, prioritizing real-time performance and resource efficiency. The proposed architecture is ideal for deployment on mobile and embedded devices since it maintains high accuracy while significantly reducing the number of trainable parameters. Our findings advance the feasibility of deploying advanced image restoration techniques in real-world, resource-constrained environments.

关键词： Depth-wise Separable Convolutions Embedded Devices Image Restoration Lightweight Real-Time Performance Resource-Constrained Platforms Deep Learning

来源：评论

学校读者我要写书评

暂无评论

Vibration analysis of elastically supported sandwich beam with damaged core and functionally graded face sheets

引用

Steel and Composite Structures 2025年第1期54卷 71-83页

作者： Wang, Lulu Mohsin, Hasan F. Bains, Pardeep Singh Sharma, Rohit Bohlooli, Nasrin College of Mechanical and Electrical Engineering Guangdong University of Science and Technology Guangdong Dongguan523083 China Computer Techniques Engineering Department Faculty of Information Technology Imam Ja’afar Al-Sadiq University Baghdad10011 Iraq Computer Technical Engineering Department College of Technical Engineering The Islamic University Najaf54001 Iraq University Karnataka Bengaluru560069 India Department of Mechanical Engineering Vivekananda Global University Rajasthan Jaipur303012 India School of Engineering and Technology Shobhit University Uttar Pradesh Gangoh247341 India Department of Mechanical Engineering Arka Jain University Jharkhand Jamshedpur831001 India Nabi Data Science & Computational Intelligence Research Co. Tehran Iran

This paper presents the influence of carbon nanotubes (CNTs) waviness, aspect ratio and damaged core on the vibrational behavior of functionally graded nanocomposite sandwich beams resting on two-parameter elastic foundations. The distributions of CNTs are considered functionally graded (FG) or uniform along the thickness of upper and bottom layers of the sandwich beam and their mechanical properties are estimated by an extended rule of mixture. In this study, the classical theory concerning the mechanical efficiency of a matrix embedding finite length fibers has been modified by introducing the tube-to-tube random contact, which explicitly accounts for the progressive reduction of the tubes’ effective aspect ratio as the filler content increases. A damage model is introduced to provide an analytical description of an irreversible rheological process that causes the decay of the mechanical properties, in terms of engineering constants. An isotropic damage is considered for the core of the sandwich beams. The equations of motion are derived based on Timoshenko beam theory and employing Hamilton’s principle. The problem is modeled using a semi-analytical approach composed of generalized differential quadrature method (GDQM) and series solution adopted to solve the equations of motion. Detailed parametric studies are carried out to investigate the mechanical behavior of these multi-layered structures depending on the damage features, through-the-thickness distribution and boundary conditions. Also, the effects of carbon nanotubes (CNTs) waviness, CNT aspect ratio, Winkler foundation modulus, shear elastic foundation modulus and geometrical conditions on the vibrational behavior of the sandwich structure are studied. © 2025 Techno-Press, Ltd.

关键词： Aspect ratio

来源：评论

学校读者我要写书评

暂无评论

Design and Implementation of a MPPT Controller-Based Bidirectional DC-DC Converter for Electric Vehicles Using PUMA Algorithm 4

Design and Implementation of a MPPT Controller-Based Bidirec...

引用

4th International Conference on Sentiment Analysis and Deep Learning, ICSADL 2025

作者： Subramaniam, G. Devayani, G. Gokulapriya, C. Janani, S. Dhanush, T. Shashhwath, S. M. Kumarasamy College of Engineering Electrical and Electronics Engineering Tamil Nadu Karur India M. Kumarasamy College of Engineering Computer Science and Business Systems Tamil Nadu Karur India

ISBN: (数字)9798331523923

ISBN: (纸本)9798331523923

The growing need for effective energy management in electric vehicles (EVs) has led to advancements in power conversion technologies, especially in systems that facilitate the use of renewable energy sources and efficient battery management. The design and implementation of bidirectional DC-DC converters in electric cars utilize Maximum Power Point Tracking (MPPT) for their controllers. This approach ensures optimal energy conversion through the PUMA (Perturb and Observe with Modified Algorithm) technique. By intelligently regulating the power flow between a renewable energy source and the vehicle's energy storage system (such as solar panels and batteries), the proposed converter maximizes energy conversion from the renewable source while keeping operational costs low and maintaining ideal battery conditions. A bidirectional converter enables the battery to both charge and discharge, while also providing a seamless interface for energy exchange between the grid, solar panels, and the car. The MPPT controller utilizes the PUMA algorithm, which is known for its faster convergence and greater stability, making it superior to traditional Disturb and Observe (P&O) algorithms. This device includes various components that monitor voltage and current values in real-time, employing adaptive control and improved power management to ensure high-quality energy conversion regardless of external conditions. Simulations and hardware implementations have shown that this method effectively enhances vehicle performance, extends the range of electric vehicles, and optimizes the use of renewable energy sources. The study also addresses several other issues, including the potential for large-scale implementation of modern electric vehicles (EVs), the robustness of the system, and the reduction of voltage and current ripple. © 2025 IEEE.

关键词： Adaptive control systems

来源：评论

学校读者我要写书评

暂无评论

Autonomous Floating Garbage Collection Device using computer Vision and Robotics 24

Autonomous Floating Garbage Collection Device using Computer...

引用

8th International Conference on Algorithms, Computing and Systems, ICACS 2024

作者： Tamgadge, Dhammashree C. Nagpure, Bhanu R. Mandal, Riddhi S. Electronics Engineering G H Raisoni College of Engineering Nagpur India Computer Science and Engineering G H Raisoni College of Engineering Nagpur Nagpur India Hong Kong Metroplotian University Kowloon Hong Kong

ISBN: (纸本)9798400718304

This paper presents an autonomous garbage collection device, which uses computer vision and robotics to clean the floating waste in the local water bodies. The tedious and increasingly repetitive task of garbage collection is becoming a major problem worldwide, our device uses ESP32-CAM and Arduino UNO to automate this task, images of water bodies are streamed through the ESP32-CAM which is then analysed by python program on the local server created by the ESP32-CAM. Arduino UNO processes the algorithm for autonomous motion of the device, currently the algorithm used is very basic, just enough to be able to make the device move autonomously, there is scope to improve the autonomous movement algorithm. we have also added an ultrasonic sensor for obstacle detection so that devices motion is not interrupted by obstacles, for garbage detection we have used YOLO-V3 algorithm which is chosen for it fast performance [5], overall we the result of our testing showed that the device is very capable of collecting floating water waste and can be deployed in local water bodies for garbage collection although there are various areas where the device can be improved one of them being use of advanced microcontroller board for local processing of images. © 2024 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： Ultrasonic sensors

来源：评论

学校读者我要写书评

暂无评论

All-Rounder: A Flexible AI Accelerator With Diverse Data Format Support and Morphable Structure for Multi-DNN Processing

引用

IEEE Transactions on Very Large Scale Integration (VLSI) Systems 2025年第5期33卷 1264-1277页

作者： Noh, Seock-Hwan Lee, Seungpyo Shin, Banseok Park, Sehun Jang, Yongjoo Kung, Jaeha Department of Electrical Engineering and Computer Science Daegu42988 Korea Republic of DGIST Daegu42988 Korea Republic of Fitogether Seoul04323 Korea Republic of Samsung Electronics Suwon16677 Korea Republic of Korea University School of Electrical Engineering Seoul02841 Korea Republic of

Recognizing the explosive increase in the use of artificial intelligence (AI)-based applications, several industrial companies developed custom application-specific integrated circuits (ASICs) (e.g., Google TPU, IBM RaPiD, and Intel NNP-I/NNP-T) and constructed a hyperscale cloud infrastructure with them. These ASICs perform operations of the inference or training process of AI models which are requested by users. Since the AI models have different data formats and types of operations, the ASICs need to support diverse data formats and various operation shapes. However, the previous ASIC solutions do not or less fulfill these requirements. To overcome these limitations, we first present an area-efficient multiplier, named all-in-one multiplier, which supports multiple bit-widths for both integer (INT) and floating-point (FP) data types. Then, we build a multiply-and-accumulation (MAC) array equipped with these multipliers with multiformat support. In addition, the MAC array can be partitioned into multiple blocks that can be flexibly fused to support various deep neural network (DNN) operation types. We evaluate the practical effectiveness of the proposed MAC array by making an accelerator out of it, named All-rounder. According to our evaluation, the proposed all-in-one multiplier occupies 1.49x smaller area compared to the baselines with dedicated multipliers for each data format. Then, we compare the performance and energy efficiency of the proposed All-rounder with three different accelerators showing consistent speedup and higher efficiency across various AI benchmarks from vision to large language model (LLM)-based language tasks. © 1993-2012 IEEE.

关键词： Application specific integrated circuits

来源：评论

学校读者我要写书评

暂无评论

Predicting Sevearity of Type 2 Diabetic Retinopathy Using Retinal Fundus Images with new DeepLearning Architectures from the Inception and Resnet Families

Predicting Sevearity of Type 2 Diabetic Retinopathy Using Re...

引用

Artificial Intelligence and Data engineering (AIDE), International Conference on

作者： Rishitha Nandanapu Krishna GSVV V Sai Sruthi Reddy Computer Science Engineering Amrita Vishwa Vidyapeetham Amaravati Vijayawada India Computer Science Engineering Amrita Vishwa Vidyapeetham Amaravati Bangaluru India Computer Science Engineering Amrita Vishwa Vidyapeetham Amaravati Hyderabad India

ISBN: (数字)9798331527518

ISBN: (纸本)9798331527525

Due to its high morbidity and medical expenses, DR mellitus (T2DM) is a serious global health concern. Its burden can be lessened with effective early prognosis, but the available treatments are not enough. Predictive modelling of Retina Images (fundus) can enhance early diagnosis and care quality. Advanced convolutional neural networks (Inception-V4, ResNet V1, and ResNet V2) are used in this study to assess the severity of Diabetic Retinopathy caused by T2DM. Using de-identified fundus data from 9,948 individuals (1,904 with a T2DM diagnosed), we used transfer learning to train these models. We have performed simple data augmentation to rectify the class imbalance. The accuracy and AUC of the ensemble model were 94.21% and 98.53%, respectively. The findings imply that the combination of ResNet and Inception-V4 enhances classification power, providing a viable approach for early T2DM identification from fundus data that may find use in clinical settings. Future research still needs to focus on improving sensitivity and integrating real-time.

关键词： Diabetic retinopathy Accuracy Sensitivity Transfer learning Retina Real-time systems Convolutional neural networks Reliability Prognostics and health management Residual neural networks

来源：评论

学校读者我要写书评

暂无评论

MUDRA: A Multi-dimensional Privacy Attainment Framework for Healthcare Blockchain Transactions

引用

IEEE Transactions on Consumer Electronics 2025年

作者： Chhabra, Archana Saha, Rahul Kumar, Gulshan Kim, Tai Hoon Lovely Professional University School of Computer Science and Engineering Punjab India Chonnam National University School of Electrical and Computer Engineering Korea Republic of

The integration of blockchain technology into consumer electronics is transforming healthcare by enhancing data security and streamlining operations in devices like wearable health monitors, smart medical devices, and electronic health records. However, this integration raises significant privacy concerns, especially regarding potential data leaks through automated processes like smart contracts. Traditional Private Information Retrieval (PIR) methods struggle with balancing data accessibility and patient confidentiality. This creates computational inefficiencies and interoperability challenges that hinder the implementation of standardized privacy measures in blockchain. We introduce the MUlti-Dimensional pRivacy Attainment (MUDRA) framework, a novel solution addressing critical privacy challenges in the storage and retrieval of patient and healthcare service provider (HSP) data. MUDRA leverages the innovative Dual Server Ring Private Information Retrieval (DS-RPIR) protocol to securely retrieve Personal Health Information (PHI) without publishing sensitive data to the blockchain, ensuring enhanced privacy and preventing unauthorized access. By integrating the Interplanetary File System (IPFS), MUDRA preserves data confidentiality and supports time-controlled revocation, allowing patients to set time-limited access to their medical records, automatically revoking access after the specified period. The framework introduces a new metric, Privacy Attainment (PA), to quantify transaction privacy and evaluate the private retrieval of sensitive data. When implemented on the Hyperledger Fabric network, MUDRA achieves 3.5% higher throughput (59 TPS), reduced **time complexity O(nlogn), and 8 ms latency-2.25 ms less than existing solutions. These performance improvements, coupled with robust privacy features, position MUDRA as a pioneering solution for addressing privacy concerns in the rapidly evolving field of healthcare consumer electronics. © 2025 IEEE.

关键词： Electronic health record

来源：评论

学校读者我要写书评

暂无评论

Comprehensive Review of Physiological Signal-Based Emotion Recognition: Methods, Challenges, and Insights on Arousal and Valence Dimensions

引用

Procedia computer science 2025年 257卷 174-181页

作者： Yassir Kacimi Mehdi Adda Mathematics Computer Science and Engineering department University of Quebec at Rimouski Canada (QC)

This review examines emotion recognition from physiological signals, focusing on processing methods and classification techniques. Key aspects include signal acquisition challenges, emotion elicitation strategies, and feature extraction methods across time, frequency, and time-frequency domains. The analysis highlights the strengths of these methods in capturing emotional information from signals such as ECG, GSR, RSP, SKT, BVP, EMG, and EOG. Feature selection and dimensionality reduction are explored regarding their role in optimizing the feature space for classification. The review also evaluates machine learning approaches and their applications in emotion recognition. This work addresses current capabilities, limitations, and emerging trends and provides a comprehensive overview of physiological signal-based emotion recognition.

关键词： Emotion Recognition Physiological Signals Signal Processing Techniques Feature Extraction Selection Machine Learning for Emotion Classification

来源：评论

学校读者我要写书评

暂无评论

A Robust Machine Learning-Based Mechanism for Detection and Analysis of Malware Attacks

引用

Procedia computer science 2025年 259卷 193-201页

作者： Rabia Bakshi Sakshi Lingwal Kumud Chandra Bhatt Siddhant Thapliyal Mohammad Wazid Devesh Pratap Singh Department of Computer Science and Engineering Graphic Era Deemed To Be University Dehradun India

Cybercriminals (hackers) may exploit potential vulnerabilities in a code or system for malicious reasons. As global innovation grows, so does cybersecurity, leading attackers to become more sophisticated. The attackers (hackers) are waiting for a chance to exploit our fault in order to create a backdoor for their malicious code. In light of these obstacles, we must be more attentive and sophisticated than the adversaries. The use of smart gadgets will gradually increase. As a result, we must develop an effective plan for protecting these smart devices. Given these considerations, we present a viable technique for machine learning-based malware attack detection and analysis (abbreviated MLBM-DAM). The proposed MLBM-DAM can detect malware inside any application, file, or system. In MLBM-DAM, we used a variety of machine learning algorithms to identify and analyze malware assaults, including random forest, decision tree, and k-nearest neighbors. The suggested MLBM-DAM’s practical implementation is described, as well as critical performance measures like as precision, recall, and accuracy. In addition, the suggested MLBM-DAM’s performance is compared to that of several competing approaches. The exhibited MLBM-DAM outperformed its current rivals.

关键词： Cyber Security Malware Distributed Denial-of-Service (DDoS) Attacks Machine learning Simulation

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：