检索结果-内蒙古大学图书馆

15th International Conference on Advances in Computing, Control, and Telecommunication Technologies, ACT 2024

作者： Ghadage, Dipali Indrajeet Gavali, Harshal Prabhakar Thodupunooru, Adharva Kumar Harikrishna Powar, Gourav Tukaram Chinchkar, Rohan Dnyandeo Computer Science & Engineering Rajarambapu Institute of Technology Sakhrale India

ISBN: (纸本)9798331300579

This research paper addresses the enhancement of data security and privacy in cloud storage through diverse encryption methods, such as one-to-many encryption, data integrity, resilient data deletion, and privacy-preserving solutions. Leveraging technologies like Advanced Encryption Standards (AES), Rivest Shamir Adleman (RSA) and searchable encryption, the study integrates privacy-preserving techniques and machine learning in cloud environments, including the use of a load balancer with GitHub to optimize data distribution and storage management for a single user across multiple repositories. Additionally, the research delves into post-quantum encryption to fortify security measures against emerging threats, underscoring the ongoing need for exploring evolving data encryption technologies in cloud storage. The paper concludes by emphasizing the significance of continuous research aligned with identified security needs, with future plans to delve deeper into evolving security requirements and the role of load balancing in optimizing data storage and distribution. © Grenze Scientific Society, 2024.

关键词： Cloud platforms

来源：评论

学校读者我要写书评

暂无评论

Simplifying Image Captioning in Hindi with Deep Learning

Simplifying Image Captioning in Hindi with Deep Learning

引用

2024 International Conference on computer, Electronics, Electrical engineering and their Applications, IC2E3 2024

作者： Rawat, Shimpi Manwal, Manika Purohit, Kamlesh Chandra Graphic Era Hill University Computer Science and Engineering Department Dehradun India Graphic Era Deemed to be University Computer Science and Engineering Department Dehradun India

ISBN: (纸本)9798350388534

We aim to simplify and optimize the process of generating Hindi captions for visual content using deep learning models. With the advancements in deep learning, generating textual descriptions for images is now possible. Any spelling, grammar, and punctuation errors have been corrected. However, extending this capability to languages other than English presents notable challenges. To bridge this gap, we propose a streamlined approach for generating image captions in Hindi. Our proposed methodology utilizes a neural network model that employs a multilayered CNN-LSTM architecture with attention mechanisms to enhance caption generation. The model has been trained on a dataset that has been sourced from MSCOCO and we have adapted and optimized hyper parameters to increase the probability of generating accurate Hindi descriptions. To evaluate the effectiveness of our models, we use BLEU scores that demonstrate significant improvements compared to existing work in this domain. Our research holds great significance beyond just technological innovation. We acknowledge the potential impact of automatic description of events in the environment and their translation into captions or messages on society. By simplifying the image captioning process in Hindi, we aim to promote linguistic inclusivity and provide a convenient interface for Hindi speakers, encouraging them to engage more easily with the diverse digital landscape. As we embark on this journey, our dedication goes beyond the intricacies of deep learning. Our goal is to simplify and make the process of generating image captions accessible and effective for the Hindi-speaking community. This research is in line with the wider discussion on linguistic diversity, cultural relevance, and the democratization of digital experiences in the ever-changing landscape of language technology. © 2024 IEEE.

关键词： Multilayer neural networks

来源：评论

学校读者我要写书评

暂无评论

Hardware-friendly rate estimation algorithm and architecture design for AVS3

引用

Multimedia Tools and Applications 2025年 1-19页

作者： Yan, Yunyao Xiang, Guoqing Chen, Jie Huang, Xiaofeng Zhang, Peng Jia, Huizhu Xie, Xiaodong School of Electronic and Computer Engineering Peking University Shenzhen China School of Computer Science Peking University Beijing China Advanced Institute of Information Technology Peking University Hangzhou China School of Communication Engineering Hangzhou Dianzi University Hangzhou China

To enhance the capabilities of advanced video coding for emerging applications, the AVS3 standard has been introduced to double the coding efficiency compared to its predecessor, the AVS2 standard. It incorporates sophisticated coding tools, such as advanced rate-distortion optimization (RDO), to select the optimal coding mode. The AVS3 standard further refines bit rate compression through Advanced Entropy Coding (AEC) within the RDO process. However, AEC demands significant computational resources due to its high data dependency and involves complex steps like binarization, context modeling, interval subdivision, renormalization, handling of outstanding bits, and context updating. Notably, the time complexity of AEC is predominantly due to context updating and interval subdivision, which are recursive and thus pose challenges for real-time applications, particularly in hardware implementations. Addressing these challenges, this paper introduces an adaptive rate estimation algorithm using a hardware-friendly piece-wise linear function to expedite bit rate calculations for AVS3 applications. Initially, we develop a linear rate estimation method based on the statistics of bin numbers. To preserve high coding performance, we further propose an adaptive model parameter updating method that adjusts weights dynamically based on varying video content. Building on this fast rate estimation algorithm, we propose a high-throughput hardware architecture that incorporates pipelining and parallelization to meet the stringent real-time requirements for 4K@120fps ultra-high-definition video at 200 MHz, with only a minimal increase in BD-rate by 1.06% under the All-Intra (AI) configuration. To our knowledge, this is the pioneering study focusing on hardware-oriented rate estimation for the AVS3 standard, marking a significant advancement in the field of video coding technology. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature

关键词： Video signal processing

来源：评论

学校读者我要写书评

暂无评论

Virtual Twin Analytics: Advancing IoT Security through AI-Powered Software Informatics 1

Virtual Twin Analytics: Advancing IoT Security through AI-Po...

引用

1st International Conference on Software, Systems and Information Technology, SSITCON 2024

作者： Nanthini, S. Jain, Upma Arora, Renuka Bhatia, Rishu Sutaria, Kruti Patil, Harshal Saveetha School of Engineering Department of Computer Science and Engineering Chennai India Graphic Era Deemed to Be University Department of Computer Science and Engineering Dehradun India Jagannath University Department of Computer Science and Engineering Haryana India Ganga Institute of Technology and Management Department of Electronics and Communication Engineering Haryana India Parul Institute of Engineering and Technology Faculty of Engineering and Technology Gujarat India School of Computer Science and Engineering Uttar Pradesh India

ISBN: (纸本)9798350352931

The Internet has grown in importance and impact over the years, causing people to become more reliant on it. The Internet has evolved into a major vector for cybercrime because to its ever-increasing user base. Over the last decade, the number of these computing systems - including desktops, laptops, smartphones, and the Internet of Things (IoT) - has skyrocketed. Among them, cell phones are practically integral to modern life. The popularity of web-based assaults has skyrocketed with the exponential growth in the number of individuals using the Internet. These web-based assaults are increasingly being combatted by security corporations. Unfortunately, new forms of these assaults are appearing all the time, making it hard for older security measures to stay up. Artificial intelligence (AI) is a source of optimism in the current cybersecurity landscape, offering a potential solution to the ever-changing digital dangers. The fast development of AI over the last decade has given rise to this optimism, because it is now impacting the expansion of every industry. With AI bringing so many advantages in every field, online security is one sector that just cannot afford to ignore it. This planned effort's work represents an advance in that direction. Critical online security issues have been the focus of this proposed work's study, which aims to address these issues using AI. Web security issues for desktop and mobile devices have been addressed in the proposed work. The planned work's contributions to online security are as follows: The 'MalCrawler' web crawler is a targeted tool for finding and exploring the web. This crawler makes it easy to gather websites, particularly ones that are harmful. It does a better job of collecting dangerous websites than a typical crawler. Additionally, it is built to circumvent the evasion strategies used by rogue websites. The crawler's ability to gather webpages - particularly dangerous ones - in order to provide datasets for ML-based an

关键词： Web crawler

来源：评论

学校读者我要写书评

暂无评论

Natural language requirements testability measurement based on requirement smells

引用

Neural Computing and Applications 2024年第21期36卷 13051-13085页

作者： Zakeri-Nasrabadi, Morteza Parsa, Saeed School of Computer Engineering Iran University of Science and Technology Hengam St. Resalat Square Tehran Tehran Iran

Requirements form the basis for defining software systems’ obligations and tasks. Testable requirements help prevent failures, reduce maintenance costs, and make it easier to perform acceptance tests. However, despite the importance of measuring and quantifying requirements testability, no automatic approach for measuring requirements testability has been proposed based on the requirements smells, which are at odds with the requirements testability. This paper presents a mathematical model to evaluate and rank the natural language requirements testability based on an extensive set of nine requirements smells, detected automatically, and acceptance test efforts determined by requirement length and its application domain. Most of the smells stem from uncountable adjectives, context-sensitive, and ambiguous words. A comprehensive dictionary is required to detect such words. We offer a neural word embedding technique to generate such a dictionary automatically. Using the dictionary, we could automatically detect Polysemy smell (domain-specific ambiguity) for the first time in 10 application domains. Our empirical study on nearly 1000 software requirements from six well-known industrial and academic projects demonstrates that the proposed smell detection approach outperforms Smella, a state-of-the-art tool, in detecting requirements smells. The Precision and Recall of smell detection are improved with an average of 0.03 and 0.33, respectively, compared to the state-of-the-art. The proposed requirement testability model measures the testability of 985 requirements with a mean absolute error of 0.12 and a mean squared error of 0.03, demonstrating the model’s potential for practical use. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

Hybrid Optimisation with Black Hole Algorithm for Improving Network Lifespan

引用

Intelligent Automation & Soft Computing 2023年第2期35卷 1873-1887页

作者： S.Siamala Devi Chandrakala Kuruba Yunyoung Nam Mohamed Abouhawwash Department of Computer Science and Engineering Sri Krishna College of TechnologyCoimbatore641042India Department of Computer Science and Engineering Vignan Nirula Institute of Technology and ScienceGuntur522009India Department of Computer Science and Engineering Soonchunhyang UniversityAsan31538Korea Department of Mathematics Faculty of ScienceMansoura UniversityMansoura35516Egypt Department of Computational Mathematics Scienceand Engineering(CMSE)Michigan State UniversityEast LansingMI48824USA

Wireless sensor networks(WSNs)are projected to have a wide range of applications in the *** fundamental problem with WSN is that it has aﬁnite *** a network is a common strategy for increasing the life-time of WSNs and,as a result,allowing for faster data *** cluster-ing algorithm’s goal is to select the best cluster head(CH).In the existing system,Hybrid grey wolf sunﬂower optimization algorithm(HGWSFO)and optimal clus-ter head selection method is *** does not provide better competence and out-put in the ***,the proposed Hybrid Grey Wolf Ant Colony Optimisation(HGWACO)algorithm is used for reducing the energy utilization and enhances the lifespan of the *** hole method is used for selecting the cluster heads(CHs).The ant colony optimization(ACO)technique is used toﬁnd the route among origin CH and *** open cache of nodes,trans-mission power,and proximity are used to improve the CH *** grey wolf optimisation(GWO)technique is the most recent and well-known optimiser module which deals with grey wolves’hunting activity(GWs).These GWs have the ability to track down and encircle *** GWO method was inspired by this hunting *** proposed HGWACO improves the duration of the net-work,minimizes the power consumption,also it works with the large-scale *** HGWACO method achieves 25.64%of residual energy,25.64%of alive nodes,40.65%of dead nodes also it enhances the lifetime of the network.

关键词： Energy efﬁciency power consumption lifespan of the network black hole method ant colony optimisation routing and cluster heads(CHs)

来源：评论

学校读者我要写书评

暂无评论

EDUVERSE: Inclusive and Immersive Metaverse Learning Environment for Education

EDUVERSE: Inclusive and Immersive Metaverse Learning Environ...

引用

2025 IEEE International Students' Conference on Electrical, Electronics and computer science, SCEECS 2025

作者： Madhubala, R. Gomathi, V. Jothikumar, Karthikeyan National Engineering College Department of Computer Science and Engineering Tamil Nadu Kovilpatti India

ISBN: (纸本)9798331529833

Eduverse is an advanced Virtual Reality (VR) platform developed to transform traditional education by creating interactive and immersive virtual learning environments. Designed to support both students and teachers, Eduverse simulates real-world classrooms, discussion rooms, and conference halls, enabling real-time communication and collaboration. One of its standout features is the accessibility tools, such as text-to-speech (TTS) and speech-to-text (STT) systems, which make the platform highly beneficial for students with disabilities. The TTS system helps students with speech impairments by converting their typed messages into spoken words, while the STT system transcribes spoken communication into text for students who have difficulty *** addition to accessibility, Eduverse offers customizable avatars, allowing students to represent themselves digitally, which helps reduce social anxiety and encourages participation. The platform also provides teachers with speech management tools, enabling them to moderate discussions, assign speaking privileges, and manage class activities. This ensures that every student has the opportunity to contribute, even in larger classes. Eduverse has been tested within a set of students, showing significant improvements in student engagement, participation, and inclusivity compared to traditional online learning platforms. Students with disabilities reported that the integrated accessibility tools allowed them to engage more fully in class discussions, making Eduverse a valuable solution for inclusive education. By combining a collaborative and supportive learning space, Eduverse demonstrates the potential to improve the way students learn and interact in the virtual world. © 2025 IEEE.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

Supervised Learning-Driven Emotion Recognition through Facial Landmark Trajectories Enhanced by SMOTE Methodology

Supervised Learning-Driven Emotion Recognition through Facia...

引用

2025 AI-Driven Smart Healthcare for Society 5.0, AdSoc5.0 2025

作者： Kar, Ishita Chatterjee, Mehuli Manna, Sukanya Das, Akash Dutta, Stobak Department of Computer Science and Engineering St Thomas’ College of Engineering and Technology Kolkata India

ISBN: (纸本)9798331536336

The study highlights the importance of facial landmark trajectories in understanding mental health by investigating the application of supervised learning systems to identify emotions. Diagnoses and treatment plans are impacted by the difficulty in recognizing and controlling emotions, which are intrinsically tied to mental health. Although physiological signals like ECG and EEG are accurate, they have drawbacks including expense and user discomfort. Because they are noninvasive, facial landmarks present a viable substitute for practical uses. Using information from the ASCERTAIN dataset, this study divided emotions into four valence-arousal plane quadrants: High Arousal High Valence (HAHV), High Arousal Low Valence (HALV), Low Arousal High Valence (LAHV), and Low Arousal Low Valence (LALV). Algorithms including Random Forest, K-Nearest Neighbors (KNN), Decision Tree, and Support Vector Machine (SVM) were used to solve a 4-class classification problem. A common problem, class imbalance, was lessened by employing the Synthetic Minority Oversampling Technique (SMOTE). The study showed that SMOTE greatly enhanced predictions for minority classes, particularly for LALV, whereas classifiers strongly preferred the majority class in the absence of balance. Despite possible overfitting, Decision Tree has the greatest true positive rate among classifiers for LALV after SMOTE application. The findings demonstrate developments in emotional computing and provide information for applications in mental health support and human-computer interaction. © 2025 IEEE.

关键词： k-nearest neighbors

来源：评论

学校读者我要写书评

暂无评论

Numerical Analysis of Pareto Set Modeling 13th

Numerical Analysis of Pareto Set Modeling

引用

13th International Conference on Evolutionary Multi-Criterion Optimization, EMO 2025

作者： Shu, Tianye Ishibuchi, Hisao Nan, Yang Pang, Lie Meng Department of Computer Science and Engineering Southern University of Science and Technology Shenzhen518055 China

ISBN: (纸本)9789819635375

Recently, neural network-based inverse models have been used for multi-objective optimization. The basic idea is to approximate the mapping from the Pareto front to the Pareto set. In general, inverse modeling from a low-dimensional objective space to a high-dimensional decision space is difficult. However, good results have been reported in the literature. In this paper, we numerically examine the performance of inverse modeling. For visual examination, we use 2- and 3-objective distance minimization problems with 2 to 1,000 decision variables. Our experimental results show that inverse modeling improves the quality of the final population, which is used as the training data. One interesting observation is that improving fitting accuracy by increasing the neural network complexity does not improve the quality of the model outputs. We also demonstrate that inverse modeling works well even for large-scale multi-objective problems with 1,000 decision variables when high-quality training data are available. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Neural network models

来源：评论

学校读者我要写书评

暂无评论

Employing Vision Transformer for Monocular Depth Estimation 2

Employing Vision Transformer for Monocular Depth Estimation

引用

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

作者： Chaudhary, Anubhav Tripathi, Dev Brat Mishra, Harsh Pathak, Prasoon Dixit, Ashish Department of Computer Science and Engineering Ajay Kumar Garg Engineering College Ghaziabad India

ISBN: (纸本)9798331530389

Monocular depth estimation (MDE) is an important task in computer vision, it enables a range of applications like robotic navigation, augmented reality and also used in surgical guidance. This paper shows the use of ViTs for MDE and focuses on the Depth Anything V2(DA-V2) model. DAV2 model leverages a teacher-student training approach and uses the synthetic and unlabeled real images which are designed to enhance the accuracy of depth estimation. DA-V2 is fine-tuned on EndoSLAM dataset and contains challenging visual scenes from endoscopic images of the stomach, colon and small intestine to check the adaptability of the model to medical depth estimation. Our experiment shows that the fine-tuned DA-V2 outperforms all the conventional models like Endo-SfMLearner and achieves reductions in root mean squared error(RMSE) and absolute relative error, across all anatomical datasets. DA-V2 achieves the RMSE values 1.245, 1.028, and 2.021 for the stomach, colon, and intestine datasets, respectively as compared to the SfMLearner's 6.387, 6.657, and 14.787. The findings show that ViTs are capable of achieving higher accuracy and generalization in monocular depth estimation tasks. This study highlights the effectiveness of DA-V2 and Vision Transformers in Monocular Depth Estimation and opens up possibilities for further model optimization. © 2025 IEEE.

关键词： Convolutional neural networks

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：