检索结果-内蒙古大学图书馆

Smart and Sustainable Developments in Electrical engineering (SSDEE), IEEE International Conference on

作者： Madhuri Gupta Megha Mishra Jaspal Bagga Department of Computer Science Shri Shankaracharya Technical Campus Rungta College of Engineering & Technology(Work) Bhilai India Department of Computer Science Shri Shankaracharya Technical Campus Bhilai India Department of Information Technology Shri Shankaracharya Technical Campus Bhilai India

ISBN: (数字)9798331542108

ISBN: (纸本)9798331542115

Chhattisgarhi, spoken by over 12 million people in Chhattisgarh, is the most prominent, holding official language status and serving as a cultural and linguistic identifier. Baheliya and Chamari, though less documented, are vital to the region's linguistic diversity. This research investigates the acoustic and phonetic characteristics of three central Indian dialects—Chhattisgarhi, Baheliya, and Chamari—using advanced machine learning techniques and robust feature extraction methods. Leveraging Mel-Frequency Cepstral Coefficients (MFCCs) for feature extraction, this study examines statistical parameters such as mean, median, and variance to distinguish the spectral properties of these dialects. Our contribution includes a detailed analysis of dialectal variations, revealing that Chamari exhibits the highest mean MFCC1 value (0.223), indicative of its higher spectral energy, while Chhattisgarhi shows the most negative mean MFCC2 (-0.322), reflecting unique spectral distinctions. Variance analysis highlights Baheliya and Chhattisgarhi as having greater spectral diversity, particularly in MFCC1 and MFCC2, showcasing their phonetic variability. These findings underline the acoustic uniqueness of each dialect, bridging gaps in existing research on underrepresented dialects. This work contributes to the preservation of linguistic diversity and provides a strong foundation for designing ASR systems tailored to regional dialects, addressing critical challenges in multilingual speech recognition.

关键词： Measurement Electrical engineering Visualization Machine learning Phonetics Signal processing Feature extraction Multilingual Mel frequency cepstral coefficient Sustainable development

来源：评论

学校读者我要写书评

暂无评论

Hybrid optimization driven fake news detection using reinforced transformer models

引用

Scientific Reports 2025年第1期15卷 1-16页

作者： M, Ganesh Karthik Faizz Ahmad, Khadri Syed Pamidimukkala, Sai Geetha Sathe, Asha Prashant G.N.V.G, Sirisha M, Sitha Ram Ch, Koteswararao Department of Computer Science and Engineering GITAM School of Technology GITAM University- Bengaluru Campus Bengaluru India Department of Computer Science and Engineering Acharya Nagarjuna University Guntur India Department of Computer Science and Engineering Koneru Lakshmaiah Education Foundation Vaddeswaram India Department of Computer Engineering Army Institute of Technology Pune India Department of CSE S.R.K.R. Engineering College Bhimavaram India Department of Computer Science and Engineering SRM University - AP Andhra Pradesh Mangalagiri India Department of Information Technology Guru Ghasidas Vishwavidyalaya Bilaspur (CG) 495009 India

The large-scale production of multimodal fake news, combining text and images, presents significant detection challenges due to distribution discrepancies. Traditional detectors struggle with open-world scenarios, while Large Vision-Language Models (LVLMs) lack specificity in identifying local forgeries. Existing methods often overestimate public opinion’s impact, failing to curb misinformation at early stages. This study introduces a Modified Transformer (MT) model, fine-tuned in three stages using fabricated news articles. The model is further optimized using PSODO, a hybrid Particle Swarm Optimization and Dandelion Optimization algorithm, addressing limitations such as slow convergence and local optima entrapment. PSODO enhances search efficiency by integrating global and local search strategies. Experimental results on benchmark datasets demonstrate that the proposed approach significantly improves fake news detection accuracy. The model effectively captures distribution inconsistencies and multimodal forgery details, outperforming conventional detectors and LVLMs. This research highlights the importance of integrating transformers and hybrid optimization to develop generalized, scalable, and accurate fake news detection systems. © The Author(s) 2025.

关键词： Fake news detection Hybrid optimization Large Vision-Language models Modified transformer Public opinion analysis

来源：评论

学校读者我要写书评

暂无评论

Drivable Area Detection Method in Dark Unstructured-Roads Based on CNN Data Fusion With Surface Normal Estimation

引用

IEEE Transactions on Intelligent Transportation Systems 2025年第6期26卷 8694-8706页

作者： Xue, Pengyu Pi, Dawei Wang, Hongliang Cheng, Yuejun Yan, Yongjun Sun, Xiaowang Liu, Yibo Wang, Xianhui Fan, Dingge Li, Xian Hu, Yibo Nanjing University of Science and Technology School of Mechanical Engineering Nanjing210094 China University of California at Riverside Department of Electrical and Computer Engineering RiversideCA92521 United States

Vehicle perception of roads is a critical foundation for intelligent transportation systems. While most existing research on intelligent transportation focuses on structured roads, the identification of drivable areas in complex dark-field environments remains challenging and underexplored. To address this issue, this paper proposes a drivable area detection network based on the fusion of images and point clouds. Specifically, a convolutional neural network (CNN) architecture augmented with a surface normal estimator (SNE) model is proposed. This approach significantly reduces resolution while effectively extracting and fusing features from multiple sensors. The proposed model is trained and validated using the Off-Road Freex Detection (ORFD) dataset, demonstrating its effectiveness and accuracy through comprehensive performance metrics. Finally, real-world vehicle tests are conducted to evaluate the drivable area detection system. These tests involve real-time data collection, image enhancement, and dense upsampling of point clouds. The processed multimodal data are fed into the detection network in real time for synchronous training and learning. The detection and recognition of unstructured drivable areas in a real dark environment perception platform confirm the efficacy of the proposed method. © 2000-2011 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Q-ACRY: Efficient Q-learning based Adaptive Crop Rotation technique for maximizing Yield

引用

Expert Systems with Applications 2025年 283卷

作者： Mohanty, Manas Kumar Kar, Samarjit Thakurta, Parag Kumar Guha Department of Computer Science and Engineering National Institute of Technology Durgapur Durgapur India Department of Mathematics National Institute of Technology Durgapur Durgapur India Department of Graphical Systems Vilnius Gediminas Technical University Vilnius Lithuania

This research paper addresses the challenge of maximizing crop yield through an intelligent approach in crop rotation using Q-learning and rule-based decision-making. The objective is to develop an adaptable crop rotation system that continuously monitors and refines the crop rotation sequence for enhanced efficacy and sustainability. Traditional crop rotation methods based on expert knowledge often fail to optimize crop yields, necessitating the use of computational models. The proposed approach integrates the Markov decision process (MDP) and Q-learning to determine an optimal crop rotation plan. A Q-learning approach is employed here to iteratively update Q-values and identify the best sequence of crops to be planted. The rewards are estimated based on yield values, providing a realistic assessment of the impact of crop rotation on crop yield. The effectiveness of the rotation sequence is evaluated over 50 years through year-wise yield assessments. Additionally, a rule-based decision-making process is introduced to modify the rotation sequence based on yield evaluations, ensuring continuous improvement in crop yield. The model is assessed using reward, Q-value, and yield, revealing improvements of 36.13%, 2.04%, and 35.73% in respective metrics. Simulation results confirm the proposed approach's superiority over existing methods. © 2025 Elsevier Ltd

关键词： Rotation rate

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive Analysis of Blockchain-Based Cryptographic Framework for Network Security 3

A Comprehensive Analysis of Blockchain-Based Cryptographic F...

引用

3rd IEEE International Conference on Automation and Computation, AUTOCOM 2025

作者： Vyshnavi, A. Bharathi, B. Singh, Amanveer Aggarwal, Prateek Yadav, Rakesh Kumar Karnataka Bangalore India Sathyabama Institute of Science and Technology Department of Computer Science and Engineering Chennai India Chitkara University Centre of Research Impact and Outcome Punjab Rajpura140417 India Chitkara University Chitkara Centre for Research and Development Himachal Pradesh 174103 India Maharishi School of Engineering & Technology Maharishi University of Information Technology Uttar Pradesh India

ISBN: (纸本)9798331542375

Blockchain technology is increasingly developing breakthroughs in secure and decentralized data storage. More recently, it has also garnered interest as a possible solution for network security. In particular, this paper conducts an indepth analysis of a network security cryptographic framework using Blockchain. The framework proposed again adopts the principles of blockchain technology, such as decentralization, immutability and consensus, to build a secure data environment for the network. It uses various cryptographic methods, digital signatures and Encryption and hash functions, to guarantee integrity, confidentiality, and authentication. The framework uses smart contracts to automatically execute security checks that can be enforced with predefined and approved methods. In this analysis, we aim to investigate the proposed scheme's effectiveness, scalability, and practicability in typical network security applications. It will also cover the challenges and future research directions in applying this framework in practical scenarios. This analysis will form the basis for a strong cryptographic system on the Blockchain to provide security reinforcement in networks. © 2025 IEEE.

关键词： Network security

来源：评论

学校读者我要写书评

暂无评论

Optimal Scheduling of Microgrid with Electric Vehicle Integration in Smart Grid using Progressive Graph Convolutional Network

Optimal Scheduling of Microgrid with Electric Vehicle Integr...

引用

Trends in Material science and Inventive Materials (ICTMIM), International Conference on

作者： Cuddapah Anitha Naveena Kumar R R Swapnil Uttamrao Deokar Harshal Shah Praful V. Nandankar Natrayan L Department of Computer Science and Engineering School of Computing Mohan Babu University Tirupati Andhra Pradesh Department of Mechanical Engineering SJB Institute of Technology Bengaluru Karnataka Departement of Mechanical Engineering Smt. Kashibai Navale College of Engineering Pune Computer Science and Engineering Faculty of Engineering and Technology Parul Institute of Technology Parul University Gujarat Department of Electrical Engineering Government College of Engineering Nagpur Maharashtra India Department of Mechanical Engineering Saveetha School of Engineering SIMATS Chennai Tamil Nadu India

ISBN: (数字)9798331501488

ISBN: (纸本)9798331501495

Electric vehicles (EVs) integrated into microgrids (MGs) provide multiple benefits but create complex optimal scheduling problems which emerge from energetic unpredictability and fluctuating renewable energy sources (RESs) as well as dissimilar EV charging behaviors. Standard scheduling approaches show poor results when trying to handle complex energy systems because they do not apply effective adaptations to changes. The paper introduces a Progressive Graph Convolutional Network (PGCN) to perform optimal scheduling for MGs with embedded EVs. The proposed method uses Levelized Cost of Energy minimization (LCOE) as its main objective and increases grid reliability performance. The PGCN model delivers predictions of energy needs together with renewable power output and EV charging activity patterns and electric grid distribution patterns to enable better planning decisions. By then, the proposed method is implemented on MATLAB platform and compared with various existing methods such as Hybrid Genetic Ant Colony (HGAC), Artificial Neural Network and Particle Swarm Optimization (ANN-PSO), Opposition-based Competitive Swarm Optimizer (OCSO), Gray Wolf Algorithm (GWA), and Genetic Algorithm (GA).With a LCOE of 0.048 $/kWh the proposed PGCN method demonstrates superior cost-effectiveness as it beats all other methods for energy scheduling and management.

关键词： Renewable energy sources Graph convolutional networks Optimal scheduling Microgrids Electric vehicle charging Smart grids Vehicle dynamics Standards Genetic algorithms Testing

来源：评论

学校读者我要写书评

暂无评论

Challenges and Innovations in Multimedia and Real-Time Networking: A Review of Modeling Approaches

引用

Procedia computer science 2025年 252卷 53-62页

作者： Milind Shah Kajal Parmar Priyanka Padhiyar Naina Parmar Monali Parikh Dhruvansh Patni Assistant Professor Department of Computer Engineering Sardar Vallabhbhai Patel Institute of Technology (S.V.I.T) Vasad Gujarat India Department of Computer Engineering Devang Patel Institute of Advance Technology & Research (DEPSTAR) Faculty of Technology & Engineering (FTE) Charotar University of Science & Technology (CHARUSAT) Changa 388421 India Assistant Professor Department of Information Technology Krishna School of Emerging Technology & Applied Research - Drs. Kiran & Pallavi Patel Global University (KPGU) Vadodara Gujarat India

Over the past decade, the rise of broadband and mobile Internet access has led to the widespread adoption of real-time networking and multimedia applications. These platforms have become essential for connecting individuals and supporting businesses, particularly with the growing trend of remote work. The demand for video streaming services has surged significantly and is expected to continue rising. However, this increased traffic has strained network performance, leading to congestion and diminishing the quality of service, especially during peak evening hours and the COVID-19 pandemic. Besides upgrading network infrastructure, it is crucial to develop intelligent streaming systems that adapt to network conditions and user expectations to enhance customer satisfaction. This review paper explores into the basics and introduction of Multimedia and Real-Time Networking. It also explores the advancements in Adaptive and Intelligent Networking for real-time multimedia communication. The paper formulates research questions based on the discussed topics, such as the distinctions between real-time networking and traditional networking, the ways multimedia applications adjust to fluctuating network conditions, and the challenges and limitations associated with these technologies.

关键词： Real Time Networking Multimedia Challenges Innovations Intelligent Networking

来源：评论

学校读者我要写书评

暂无评论

MM-Retinal V2: Transfer an Elite Knowledge Spark into Fundus Vision-Language Pretraining

arXiv

引用

arXiv 2025年

作者： Wu, Ruiqi Su, Na Zhang, Chenran Ma, Tengfei Zhou, Tao Cui, Zhiting Tang, Nianfeng Mao, Tianyu Zhou, Yi Fan, Wen Wu, Tianxing Jing, Shenqi Fu, Huazhu School of Computer Science and Engineering Southeast University Nanjing China Department of Ophthalmology The First Affiliated Hospital of Nanjing Medical University Nanjing China School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing China Institute of High-Performance Computing Agency for Science Technology and Research Singapore

Vision-language pretraining (VLP) has been investigated to generalize across diverse downstream tasks for fundus image analysis. Although recent methods showcase promising achievements, they significantly rely on large-scale private image-text data but pay less attention to the pretraining manner, which limits their further advancements. In this work, we introduce MM-Retinal V2, a high-quality image-text paired dataset comprising CFP, FFA, and OCT image modalities. Then, we propose a novel fundus vision-language pretraining model, namely KeepFIT V2, which is pretrained by integrating knowledge from the elite data spark into categorical public datasets. Specifically, a preliminary textual pretraining is adopted to equip the text encoder with primarily ophthalmic textual knowledge. Moreover, a hybrid image-text knowledge injection module is designed for knowledge transfer, which is essentially based on a combination of global semantic concepts from contrastive learning and local appearance details from generative learning. Extensive experiments across zero-shot, few-shot, and linear probing settings highlight the generalization and transferability of KeepFIT V2, delivering performance competitive to state-of-the-art fundus VLP models trained on large-scale private image-text datasets. Our dataset and model are publicly available via https://***/lxirich/MM-Retinal. Copyright © 2025, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

DeepCircuitX: A Comprehensive Repository-Level Dataset for RTL Code Understanding, Generation, and PPA Analysis

arXiv

引用

arXiv 2025年

作者： Li, Zeju Xu, Changran Shi, Zhengyuan Peng, Zedong Liu, Yi Zhou, Yunhao Zhou, Lingfeng Ma, Chengyu Zhong, Jianyuan Wang, Xi Zhao, Jieru Chu, Zhufei Yang, Xiaoyan Xu, Qiang Department of Computer Science and Engineering The Chinese University of Hong Kong Sha Tin Hong Kong Department of Computer Science and Engineering Shanghai China School of Computer Science Hangzhou Dianzi University Hangzhou China Faculty of Electrical Engineering and Computer Science Ningbo University Ningbo China School of Integrated Circuit Southeast University Nanjing China National Center of Technology Innovation for EDA Nanjing China

This paper introduces DeepCircuitX, a comprehensive repository-level dataset designed to advance RTL (Register Transfer Level) code understanding, generation, and power-performance-area (PPA) analysis. Unlike existing datasets that are limited to either file-level RTL code or physical layout data, DeepCircuitX provides a holistic, multilevel resource that spans repository, file, module, and block-level RTL code. This structure enables more nuanced training and evaluation of large language models (LLMs) for RTL-specific tasks. DeepCircuitX is enriched with Chain of Thought (CoT) annotations, offering detailed descriptions of functionality and structure at multiple levels. These annotations enhance its utility for a wide range of tasks, including RTL code understanding, generation, and completion. Additionally, the dataset includes synthesized netlists and PPA metrics, facilitating early-stage design exploration and enabling accurate PPA prediction directly from RTL code. We demonstrate the dataset’s effectiveness on various LLMs finetuned with our dataset and confirm the quality with human evaluations. Our results highlight DeepCircuitX as a critical resource for advancing RTL-focused machine learning applications in hardware design automation. Our data is available at https://***/lcm-team. © 2025, CC BY.

关键词： Hardware-software codesign

来源：评论

学校读者我要写书评

暂无评论

Machine learning-guided humidity-induced degradation analysis of Cs2CuSbCl6 sustainable perovskite solar cell

引用

Renewable Energy 2025年 250卷

作者： Raj, Manasvi Batra, Sajal Aggarwal, Anshul Kushwaha, Aditya Goel, Neeraj Department of Electronics and Communication Engineering Netaji Subhas University of Technology Dwarka New Delhi110078 India Department of Computer Science & Engineering SRM University Delhi-NCR Haryana Sonepat131029 India

In this study, we present a machine learning (ML)-predicted framework to evaluate both the performance and environmental stability of a novel lead-free perovskite solar cell with the structure FTO/AZnO/Cs2CuSbCl6/MoO3. Our approach integrates theoretical calculations, device simulations, and ML to predict device behaviour under varied conditions. Density Functional Theory (DFT) calculations were employed to determine intrinsic electronic properties, yielding bandgaps of 3.481 eV for FTO, 3.330 eV for AZnO, 1.7 eV for Cs2CuSbCl6, and 3.170 eV for MoO3. Optical analysis confirmed that Cs2CuSbCl6 exhibits strong UV absorption with minimal losses, positioning it as an effective absorber. Subsequently, SCAPS-1D simulations optimized photovoltaic parameters, achieving an open-circuit voltage (Voc) of 1.58 V, a short-circuit current density (Jsc) of 21.82 mA/cm2, a fill factor (FF) of 91.12 %, and a power conversion efficiency (PCE) of 31.50 %. Nearly 100 % quantum efficiency in the visible spectrum further demonstrates efficient charge carrier extraction. To address moisture-induced degradation, we developed a linear regression ML model trained on both experimental and simulated data, which achieved an R2 of 0.989. The model reliably predicts that increased humidity and prolonged exposure significantly reduce PCE. This integrated methodology of DFT, device simulation, and ML offers a powerful tool for designing robust, high-performance perovskite photovoltaics with enhanced environmental resilience. © 2025 Elsevier Ltd

关键词： Open circuit voltage

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：