检索结果-内蒙古大学图书馆

2024 IEEE International Conference on computer Vision and Machine Intelligence, CVMI 2024

作者： Naveen, Aryan Kashyap Thunga, Sunil Murki, Anuhya Kalale, Mahati Anil, Shriya National Institute of Technology Karnataka Dept. of Information Technology Surathkal India National Institute of Technology Karnataka Dept. of Computer Science and Engineering Surathkal India

ISBN: (纸本)9798350376876

With exponential growth in the use of digital image data, the need for efficient transmission methods has become imperative. Traditional image compression techniques often sacrifice image fidelity for reduced file sizes, challenging maintaining quality and efficiency. They also compromise security, leaving images vulnerable to threats such as man-in-the-middle attacks. This paper proposes an autoencoder architecture for image compression to not only help in dimensionality reduction but also inherently encrypt the images. The paper also introduces a composite loss function that combines reconstruction loss and residual loss for improved performance. The autoencoder architecture is designed to achieve optimal dimensionality reduction and regeneration accuracy while safeguarding the compressed data during transmission or storage. Images regenerated by the autoencoder are evaluated against three key metrics: reconstruction quality, compression ratio, and one-way delay during image transfer. The experiments reveal that the proposed architecture achieves an SSIM of 97.5% over the regenerated images and an average latency reduction of 87.5%, indicating its effectiveness as a secure and efficient solution for compressed image transfer. © 2024 IEEE.

关键词： Image compression

来源：评论

学校读者我要写书评

暂无评论

Accelerating Cloud Deployment Through Containerization Technologies

Accelerating Cloud Deployment Through Containerization Techn...

引用

2024 International Conference on Advances in Computing Research on Science engineering and technology, ACROSET 2024

作者： Lakkadwala, Aayushya Lakkadwala, Prashant Acropolis Institute of Technology and Research Dept. of Computer Science and Engineering Indore India Acropolis Institute of Technology and Research Dept. of Information Technology Indore India

ISBN: (纸本)9798350388800

This research investigates the effectiveness of containerization technologies, with a focus on Docker, in accelerating cloud application deployments. The study examines diverse methodologies aimed at optimizing container image size, refining container orchestration mechanisms for expedited deployments, and evaluating the integration of serverless functions within containers to reduce deployment durations. By synthesizing existing scholarly literature with empirical analyses, this paper aims to elucidate the efficacy of containerization in streamlining cloud deployment processes. It offers insights into optimal practices and identifies prospective avenues for future research and development in the field. © 2024 IEEE.

关键词： Cloud platforms

来源：评论

学校读者我要写书评

暂无评论

BanglaGPT: A Generative Pretrained Transformer-Based Model for Bangla Language 2

BanglaGPT: A Generative Pretrained Transformer-Based Model f...

引用

2nd International Conference on Information and Communication technology for Sustainable Development, ICICT4SD 2023

作者： Salim, Md. Shahidul Murad, Hasan Das, Dola Ahmed, Faisal Khulna University of Engineering Technology Dept. of Computer Science and Engineering Khulna9203 Bangladesh Chattogram University of Engineering Technology Dept. of Computer Science and Engineering Chattogram4349 Bangladesh

ISBN: (纸本)9798350358667

Natural Language Processing (NLP) has entered a new era with the advent of pre-trained language models, paving the way for constructing robust language models. Pretrained transformer-based models such as GPT-2 have become prevalent due to their cutting-edge efficiency. However, these approaches rely heavily on resource-intensive languages, forcing other languages to adopt multilingual frameworks (mGPT). The mGPT model could perform better for low-resource languages such as Bangla because the model has been trained on a diverse dataset spanning multiple languages. Recent studies show that the language-specific GPT model outperforms the multilingual mGPT model. In this research, we have proposed a pretrained monolingual GPT model called BanglaGPT using the objective of causal language modeling (CLM). Due to the lack of available large datasets for NLP tasks in Bangla, we have created a Bangla language model dataset called BanglaCLM using a 26.24 GB Bangla corpus scraped from several public websites. We have used a subword-based tokenization algorithm named Byte-Pair Encoding (BPE) for Bangla and finally trained the Bangla-GPT2 model from scratch using the BanglaCLM dataset. Our pretrained BanglaGPT provides state-of-the-art performance for Bangla text generation with a perplexity score of 2.86 and a loss score of 0.45 on the test set. © 2023 IEEE.

关键词： Large dataset

来源：评论

学校读者我要写书评

暂无评论

Multi-Agent H-Learning Based Cooperative Spectrum Sensing for Cognitive Radio Networks 32

Multi-Agent H-Learning Based Cooperative Spectrum Sensing fo...

引用

32nd International Conference on Electrical engineering, ICEE 2024

作者： Fard, E. Karimpour Nouri, M. Behroozi, H. Sobhi-Givi, S. Sharif University of Technology Dept. of Electrical Engineering Tehran Iran University of Mohaghegh Ardabili Dept. of Electrical and Computer Engineering Ardabil Iran

ISBN: (纸本)9798350376340

Fast and accurate methods for spectrum sensing (SS) are the key elements in cognitive radio networks (CRNs) that achieve high SS. This paper proposes a reinforcement learning (RL) scheme for secondary users (SUs) in a massive multiple-input multiple-output (MIMO) CRN to determine occupation and emptiness of scanned wideband channels. In the proposed H-learning method, the clustered neighboring SUs collect data to check the status of channels that improve reliability of measured sensing. Then, they share information about different frequency channels with other clusters and the status of channel occupation is determined for all SUs. This model-based learning method increases finding free channels and enhances spectral efficiency (SE) in presence of large probability of missing detection. Furthermore, each SU learns the presence pattern of primary user (PU) and gather a dynamic scan priority list to decrease scan overhead and access delays. Simulation results demonstrate the multi-agent reinforcement modified H-learning (MH-learning) method can achieve better performance in terms of PU usage percent, number of attempts, call block rate, and probability of detection in comparison to H-learning and some state-of-art learning methods. © 2024 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Robust decentralized plug and play voltage tracker of islanded microgrids under loads and lines uncertainties by the invariant ellipsoids

引用

Journal of King Saud University - engineering Sciences 2024年第6期36卷 409-419页

作者： Bayoumi, Ehab H.E. Soliman, Hisham M. El-Sheikhi, Farag Ali Shafiq, M. Awad, Hilmy Energy and Renewable Energy Department Egyptian Chinese University Cairo Egypt Dept. of Electrical Power Eng. Faculty of Eng Cairo University Cairo Egypt Dept. of Electrical and Electronics Engineering Istanbul Esenyurt University Istanbul Turkey Dept. of Electrical and Computer Engineering Sultan Qaboos University Muscat Oman Dept. of Electrical Technology Faculty of Technology and Education Helwan University Egypt

Microgrids (MGs) comprising of distributed generation units (DGs) are subjected to plug-and-play operation (PnP), lines (connecting the DGs) parameters uncertainty, and load changes. Robust stability and an authentic operation for islanded microgrids can be guaranteed through a robust decentralized voltage tracker developed in this paper. The proposed controller design has the following merits: 1) fully decentralized, 2) scalable, and 3) maintains robust stability against PnP of DGs, load changes, and lines parameters’ uncertainties. A sufficient condition is developed by linear matrix inequality convex optimization is exploited to solve the problem. The MGs’ changes in load and the line's parameters are modelled as norm-bounded uncertainties. The suggested controller uses local measurements from DGs, i.e., decentralized by the decomposition of the global system into subsystems. For each subsystem, the rest of the system's impact is considered disturbances, whose influence must be minimized. The proposed disturbance rejection control algorithm is based on the method of invariant-ellipsoids. Several time-domain scenarios such as connecting and disconnecting a number of DGs, local load changes, and variation of transmission line parameters are executed to assess the suggested controller's effectiveness, using MATLAB /Sim PowerSystems Toolbox. © 2021 The Authors

关键词： Controllers

来源：评论

学校读者我要写书评

暂无评论

Towards Developing a Transformer-Based Bangla Typing Error Correction Model: A Deep Learning-Based Approach 2

Towards Developing a Transformer-Based Bangla Typing Error C...

引用

2nd International Conference on Information and Communication technology for Sustainable Development, ICICT4SD 2023

作者： Salim, Shahidul Islam, Taqiul Zannat, Rowzatul Mia, Naeem Fuad, Mohtasim Murad, Hasan Khulna University of Engineering Technology Dept. of Computer Science and Engineering Khulna9203 Bangladesh Chattogram University of Engineering Technology Dept. of Computer Science and Engineering Chattogram4349 Bangladesh

ISBN: (纸本)9798350358667

Despite being the sixth most widely spoken language globally, the Bangla language faces a lack of effective and efficient automated methods to correct keyboard errors in written text. Such errors can result in communication misunderstandings and misinterpretations, particularly in the digital domain, where written text plays a crucial role in communication and is heavily relied upon. However, no keyboard error model for the Bangla language has been found due to the lack of available datasets. In this research work, we have proposed a probabilistic approach to generate a synthetic Bangla keyboard typing error correction (BKTEC) dataset. By utilizing the novel BKTEC dataset, we have fine-tuned a transformer-based model for Keyboard error correction. After training it to convergence, the proposed model has been tested on a test dataset, and we have found a word level accuracy of 88.56%. © 2023 IEEE.

关键词： Error correction

来源：评论

学校读者我要写书评

暂无评论

Recognition of Fingerprint Images using CNN for Cybercrime Detection System 2

Recognition of Fingerprint Images using CNN for Cybercrime D...

引用

2nd IEEE International Conference on Networks, Multimedia and Information technology, NMITCON 2024

作者： Kavitha, T. Patil, Bhimaraya Saraswathi, S. Rajarajeswari, S. Patil, Anita New Horizon College of engineering Dept. of Computer Engineering Bangalore India Sri Sivasubramaniya Nadar College of Engineering Dept. of Computer Science Engineering Chennai India Vellore Institute of Technology School of computer science and engineering Vellore India City Engineering College Dept. of Information Science Bengaluru India

ISBN: (纸本)9798350372892

Fingerprint identification plays a crucial role in crime detection, but manual analysis is time-consuming and error-prone. This research aims to develop an automated system for accurate and efficient fingerprint matching using machine learning. The proposed approach involves image acquisition, preprocessing, feature extraction, and classification. Fingerprint images are enhanced, binarized, and segmented. Minutiae (ridge endings and bifurcations) are extracted as features. A convolutional neural network (CNN) is trained on a dataset of fingerprint images for classification. By comparing algorithms, Random Forest achieved over 8 1 % accuracy, outperforming Decision Trees (nearly 75% accuracy). The trained CNN model can accurately match fingerprints from crime scenes with potential suspects, providing reliable evidence. The developed system offers law enforcement agencies an automated and accurate fingerprint identification tool, improving crime-solving efficiency and reducing manual effort. The proposed approach showcases the potential of machine learning in forensic applications. © 2024 IEEE.

关键词： Random forests

来源：评论

学校读者我要写书评

暂无评论

Microwave Marvels: Design and Computational Analysis of an Ultra-thin Metamaterial Absorber for C, X and K Band 6

Microwave Marvels: Design and Computational Analysis of an U...

引用

6th International Conference on Electrical engineering and Information and Communication technology, ICEEICT 2024

作者： Tasnim, Nowshin Rattre, Naziya Hasan Shawon, Sarowar Morshed Alam, Mohammed Shamsul Alam, Touhidul Chittagong University of Engineering & Technology Dept. of Computer Science & Engineering Chittagong4349 Bangladesh International Islamic University Chittagong Dept. of Computer Science & Engineering Chittagong4318 Bangladesh University of Science & Technology Chittagong Dept. of Computer Science & Engineering Chattogram4202 Bangladesh Universiti Kebangsaan Malaysia Bangi43600 UKM Malaysia

ISBN: (纸本)9798350385779

Metamaterial absorbers are an advancement in material science as they provide more advantages in comparison to conventional materials. Achieving miniaturization with a multilayered structure is the primary challenge here. In this study, a novel metamaterial absorber design is proposed for microwave application. The proposed design is shaped as a squad of four segments where each segment contains an octagonal split ring and a central ring. The multi-layered design of the absorber helps to secure a high absorption rate. An extensive range of research is done to determine particular measurement units, and geometric shapes with the help of CST Studio Suit 2019. Here both ground and patch are designed using copper and the separator is designed with Roger. Moreover, the thickness of the absorber is only 3.4 mm which ensures its ultra-thin and lightweight nature. With an elaborated numerical analysis, it has been shown that the proposed absorber provides 90% to 99% absorption at the C, X, and K bands. Moreover, a concise computational analysis is accomplished with FR4 lossy, Rogers RT6010LM lossy, and Rogers RT5870LM lossy as dielectric constant. The key contribution of this research is the high absorption rate, coupled with minimal thickness and simple structure. © 2024 IEEE.

关键词： Metamaterials

来源：评论

学校读者我要写书评

暂无评论

Cryptosight: Decrypting Future Trends with LSTM, PyCaret and Jinja 2.0 1

Cryptosight: Decrypting Future Trends with LSTM, PyCaret and...

引用

1st IEEE International Conference on Smart Power Control and Renewable Energy, ICSPCRE 2024

作者： Mhatre, Soham Motiram Mishra, Brajesh Kumar Raja, Ishan Kumar Goswami, Bigit Krishna Agrawal, Sneha Upadhyay, Pranjal National Institute of Technology Dept. of Electronics and Communication Engineering Nagaland Dimapur India National Institute of Technology Dept. of Mechanical Engineering Nagaland Dimapur India National Institute of Technology Dept. of Electrical & Electronics Engineering Nagaland Dimapur India National Institute of Technology Dept. of Computer Science and Engineering Nagaland Dimapur India

ISBN: (纸本)9798350377002

The primary objective of this research is to develop and evaluate combined cryptocurrency price prediction models using Long Short-Term Memory (LSTM) neural networks along with traditional time series models and machine learning techniques like Random Forest and Linear Regression. The LSTM models will capture long-term dependencies in sequential cryptocurrency price data, while the other models provide complementary predictive capabilities. This multi-model approach aims to leverage the strengths of different techniques to improve overall prediction accuracy. Additionally, this research seeks to explore the practical application of these predictions in real-world cryptocurrency trading scenarios. The study will collect historical price data and relevant news for major cryptocurrencies from reputable sources. Data preprocessing, feature engineering, model training, hyper-parameter tuning, and evaluation using metrics like RMSE, MAE, and R-squared will be performed. The predictive performance of the LSTM-based models will be compared against baseline models, with statistical significance testing. Interpretability analysis will provide insights into the factors driving price movements. Furthermore, the research will investigate integrating the predictions into trading algorithms and evaluate their efficacy through back-testing on historical data. Limitations, challenges, and future research directions in cryptocurrency price prediction will also be discussed. © 2024 IEEE.

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

WellBe: An Intelligent Elderly Care and Well-Being Monitoring System Using Deep Learning 4

WellBe: An Intelligent Elderly Care and Well-Being Monitorin...

引用

4th International Conference on Intelligent Technologies, CONIT 2024

作者： Raotole, Atharv Shirodkar, Sailee Sanjay Shukla, Rahul Sisodia, Jignesh Devadkar, Kailas Patel Institute of Technology Dept. of Information Technology Sardar Mumbai India Patel Institute of Technology Dept. of Computer Science and Engineering Sardar Mumbai India

ISBN: (纸本)9798350349900

Deep Learning refers to a subset of Machine Learning that utilizes deep neural networks to simulate the complex decision-making process of the human brain. In recent years, DL has made remarkable progress in addressing multiple unique challenges faced by most elderly individuals. While there have been significant contributions to independent systems catered for the elderly, there is the existence of an overall integrated AI platform that unifies the use of individual modules for an optimized solution. This paper introduces 'WellBe,' an innovative AI system that encompasses an alarm-based fall detection component, a facial emotion recognition (FER) system and a built-in personalized chatbot all under a user-friendly website. The novelty of the work extends beyond the aforementioned components, owing to the implementation of an optimized FER system able to handle wrinkles and changes in skin texture, thereby tailored exclusively for the elderly demographic. The system leverages a custom feature extraction module, fusing facial and wrinkle features, in addition to extending a convolutional neural network (CNN) architecture for accurate emotion classification, achieving promising performance with precision, recall, and F1-scores of 0.87, 0.91, and 0.89 for the 'Happy' (one of the predicted emotions) class, respectively. The fall detection component, on the other hand, utilizes a pre-trained YOLOv7 Pose model for the real-time identification of falls, sitting at a performance accuracy of 95.3%. Finally, the personalized chatbot is implemented using a rule-based approach enhanced with the Jaccard similarity algorithm, to facilitate prompt assistance and a sense of companionship to the end user. Overall, the integration of these components into a cohesive website offers a comprehensive solution for enhancing the quality of life for elderly individuals living alone, addressing loneliness, safety, and overall well-being. © 2024 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：