检索结果-内蒙古大学图书馆

Handwritten optical character recognition using TransRNN trained with self improved flower pollination algorithm (sI-FPA)

引用

Multimedia Tools and Applications 2024年 1-23页

作者： Devi, s. Nandhini Fatima, N. sabiyath Department of Computer Science and Engineering B.S. Abdur Rahman Crescent Institute of Science & ampTechnology Chennai India

The difficulty of successfully scanning handwritten text arises from variances in style, size, and orientation, which affect handwriting optical character recognition (OCR). This study provides a novel strategy that incorporates innovative techniques for effectively addressing these difficulties. Initially, preprocessing steps involve Gaussian blur to reduce noise and CLAHE to enhance contrast, improving the visibility of text regions. Canny Edge Detection is then utilized for segmentation, followed by text localization using the Efficient and Accurate scene Text (EAsT) algorithm. Next, features are extracted via Inception V3 to capture both low-level and high-level information. From the collected characteristics, Long short-Term Memory (LsTM) networks are used to extract chronological data. Then, with the help of the self-Improved Flower Pollination Algorithm (sI-FPA), features are selected using Optimized Principal Component Analysis (PCA). Finally, a combination of Transformer and Recurrent neural networks (RNNs) is used to perform text recognition, maximizing the benefits of each network’s advantages in processing sequential and contextual data. Experiments on MNsT dataset with validation using different optimization algorithms show the effectiveness and resilience of the suggested method in different optimization scenarios. The combination of these methods results in a sophisticated OCR system that can recognize handwritten characters with accuracy, even under difficult conditions. © The Author(s), under exclusive licence to springer science+b.s.ness Media, LLC, part of springer Nature 2024.

关键词： Principal component analysis

来源：评论

学校读者我要写书评

暂无评论

IDedupNet: A MobileNetV3-b.s.d Deep Learning Framework for Efficient Image Deduplication in Cloud Computing Environments

Informatica (Slovenia)

引用

Informatica (slovenia) 2025年第13期49卷 143-162页

作者： Mohiuddin, Mohd Hasan Tamilselvan, Latha Department of Computer Science and Engineering B.S. Abdur Rahman Crescent Institute of Science &Technology Vandalur Chennai600048 India Department of Information & Technology B.S. Abdur Rahman Crescent Institute of Science &Technology Vandalur Chennai600048 India

Image deduplication is becoming increasingly important for cloud storage infrastructures to handle the increasing amount of multimedia material. Through increased storage efficiency, effective picture deduplication may optimize resources and save expenses. It also improves performance by facilitating quicker access, utilizing less bandwidth, and enhancing data integrity. Although heuristic-b.s.d classical deduplication techniques work well in various storage infrastructures, they cannot keep up with the dynamic nature of cloud storage resources. This study presents IDedupNet, a revolutionary DL-b.s.d framework that improves infrastructure performance in cloud computing by efficiently detecting duplicate and near-duplicate photos. Our approach leverages MobileNetV3 for feature extraction and CNN-b.s.d encodings for image deduplication, enabling it to manage duplicate photos in highly dynamic contexts efficiently. Additionally, we provide a Learning-b.s.d Image Deduplication (LbID) approach that improves deduplication performance by extending the use of the IDedupNet model. Experimental evaluation demonstrates a high accuracy of 98.68% on benchmark datasets, consistently outperforming existing models. The underlying technique and deep learning framework may be easily integrated into real-time cloud storage systems to increase customer satisfaction and infrastructure efficiency. © 2025 slovene society Informatika. All rights reserved.

关键词： Cloud storage

来源：评论

学校读者我要写书评

暂无评论

simulator for swarm Feature selection Algorithms for Classification Problem 2

Simulator for Swarm Feature Selection Algorithms for Classif...

引用

2nd IEEE International Conference on Interdisciplinary Approaches in Technology and Management for social Innovation, IATMsI 2024

作者： Kulkarni, Mayuri Diwakar shahade, Makarand Deore, shailesh Khandekar, Nilesh bhavsar, Ishwari Garud, Jayesh Ahirrao, Kunal SVKM's Institute of Technology Dept. of Computer Engineering Dhule India SSVPS's B.S. Deore CoE Dept. of Computer Engineering Dhule India

ISBN: (纸本)9798350360523

swarm Intelligence (sI) algorithms address difficult optimization and decision-making issues using cues from collective behavior. These behaviors are inspired by social creatures like ants, bees, and birds. Using these creatures' behaviors algorithms have demonstrated amazing promise in several fields, such as distributed computing, robotics, and optimization. Assessing sI algorithms' performance under various conditions and scenarios is still a significant difficulty, though. We introduce a new swarm Intelligence Algorithm Performance Analysis simulator (sIAPAs) to tackle this issue. This proposed simulator introduces a novel approach to automating the process of selection for swarm Intelligence (sI) algorithms on datasets. The system empowers users to upload datasets into a specialized simulator, which then orchestrates a comparative analysis using a suite of sI algorithms. The simulator evaluates and selects the most suitable algorithm b.s.d on the dataset's characteristics and problem domain. This provides the optimal selected features by the respective algorithms b.s.d on measured parameters. © 2024 IEEE.

关键词： swarm intelligence

来源：评论

学校读者我要写书评

暂无评论

A Rob.s. Parkinson's Disease Recognition from MRI brain Images by Densely Multi-Dilated Densely Connected ResNet Model

A Robust Parkinson's Disease Recognition from MRI Brain Imag...

引用

2024 International Conference on Integrated Circuits, Communication, and Computing systems, ICIC3s 2024

作者： Arogia Victor Paul, M. shankar, sharmila B.S. Abdur Rahman Crescent Institute of Science and Technology Department of Computer Science and Engineering Chennai India

ISBN: (纸本)9798350364088

Parkinson disease (PD) is a neurological ailment that can be treated that causes excruciating discomfort in the brain as well as problems with balance and dexterity. Physicians frequently employ experimental indicators, such as motorized and non-motor symptoms b.s.d on brain magnetic resonance imaging (MRI) pictures, to diagnose Parkinson's disease (PD) in a medical stage. Nevertheless, there are disparities in inter-variability because these medical evaluations heavily rely on skilled physicians. Thus, this work proposes a sophisticated deep learning (DL) method for precise brain MRI-b.s.d PD analysis. To improve the data eminence, Extended Unsharp Mask Filtering (Ex_UMF) is first used to the brain images. The suggested Multi-dilated Densely linked ResNet model (MD_DenseRNet) is then fed the preprocessed pictures to determine if the image is PD or normal. The Python platform processes and examines the generated method, and the simulation progresses using a publicly available PPMI datab.s.. A number of evaluation metrics, including computing time (CT), sensitivity, specificity, and precision, are looked at and linked to conventional research. The created MD_DenseRNet model achieved 98.87% accuracy, 97% precision, 98.1% recall, 96.9% specificity, and 0.15 ms CT in the experimental part. © 2024 IEEE.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

Risk Prediction in Financial Transactions Using IoT big Data Analytics 5

Risk Prediction in Financial Transactions Using IoT Big Data...

引用

5th International Conference on Electronics and sustainable Communication systems, ICEsC 2024

作者： Chidambaranathan, P. Muthupriya, V. B.S. Abdur Rahman Crescent Institute of Science and Technology Department of Computer Science and Engineering Tamil Nadu Chennai India

ISBN: (纸本)9798350379945

The Internet of Things is a network of appliances and other items that can communicate and distribute data;as a result, Internet of Things are equipped with actuators, sensors, microelectronics, software, and connections. IoT devices, to put simply, share data over wired or wireless networks. Inventions of IoT have endless possibilities that can make a huge difference in the world. IoT is having the most effect on b.s.ness since has altered not just how various corporate processes are conducted, but also how the economy is managed. big data and IoT help optimize operations, reducing costs, enhancing productivity and improving lives. Globally, financial institutions and banks are coming up with fresh, creative approaches to quickly assimilate big data analytics for every step of the process in with the intention of maximizing efficiency. Numerous previous studies mostly concentrate on risk assessment along with big data analytics together in the banking industry;nevertheless, might not particularly address the difficulties posed by cross-border transactions. The purpose of the proposed study is to close this gap by focusing on the particulars and bringing up the approach of big data analytics for risk assessment in cross-border transactions. © 2024 IEEE.

关键词： Microelectronics

来源：评论

学校读者我要写书评

暂无评论

Prediction and Classification of Handwritten Digits using DRNN-WOA model with Radiomics Features 4

Prediction and Classification of Handwritten Digits using DR...

引用

4th International Conference on Power, Energy, Control and Transmission systems, ICPECTs 2024

作者： Devi, s. Nandhini Fatima, N. sabiyath B.S. Abdur Rahman Crescent Institute of Science & Technology Department of Computer Science and Engineering Tamil Nadu Chennai India

ISBN: (纸本)9798331508845

Words in handwriting are notoriously difficult to identify and distinguish because of their complex and irregular shapes. Word retrieval is a challenging and time-consuming process in low-resource scripts. To get the most out of deep learning and neural network models, it is necessary to employ large amount trained datasets and present randomness into the existing structure. Existing HDR methods rely primarily on hand-coded key points data. The wide variety of writing styles, as well as size fluctuations in the input sample, make accurate recognition and categorization of numerals to be a difficult task. To overcome these constraints, this research work introduces a Deep-Recurrent (DRNN) Neural Network model and provides an effective and efficient HDR system. before considering pre-processing to enhance prediction and classification accuracy, HDR is applied to the MNIsT and DIDA datasets. both the Convolution algorithms and the Gaussian filters are used to refine the digits. The DRNN model receives the extracted features called Radiomics as input. since the Whale Optimization Algorithm (WOA) is used to fine-tune the DRNN's parameters, the suggested type is called DRNN-WOA. Accuracy, recall, F1- score, specificity, AUC, precision, and false positive rate are some metrics used to compare the proposed DRNN-WOA model. © 2024 IEEE.

关键词： Recurrent neural networks

来源：评论

学校读者我要写书评

暂无评论

Enhancing speech Emotional Recognition through a Multi-Layer Perceptron Model 2

Enhancing Speech Emotional Recognition through a Multi-Layer...

引用

2nd International Conference on Trends in Electrical, Electronics and computer engineering, TEECCON 2023

作者： Lakshminarayanan, Arun Raj singh Rathna balaji, Inderpaal Hussain, sheik Taaha Jayaraman, Vaishnavi Anwar, Chintakunta shaik B.S. Abdur Rahman Crescent Institute of Science and Technology Department of Computer Science and Engineering Chennai India

ISBN: (纸本)9798350339949

Detecting emotions in audio speech files is a significant challenge for machines, despite the natural skill of humans in this area. While computers excel at comprehending informational content, deciphering underlying emotions remains a formidable challenge. At this juncture, the speech Emotion Recognition system (sERs) emerges as an important component of the solution. To reliably categorize audio speech data into distinct emotional categories such as happiness, sorrow, anger, and neutrality, a specific model called sERs was developed. It has a wide range of potential uses, from healthcare to contact centers. In this study, the following Machine Learning models were utilized support Vector Machine (sVM), k-nearest neighbors (KNN), and Decision Tree (DT). These models yielded lower accuracy rates when categorizing emotions. Thus, to improve the accuracy rate three crucial audio features such as Mel, chroma, and MFCC are applied by the constructed sERs and analyzed by utilizing a grid search-b.s.d Multi-Layer Perceptron (MLP) classifier for accurate results. The proposed model produced 75.97% accuracy rate, which is quite remarkable. © 2023 IEEE.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive Investigation on Real - Time Object Detection in Deep Learning 5

A Comprehensive Investigation on Real - Time Object Detectio...

引用

5th IEEE International Conference on Advances in Electronics, computers and Communications, ICAECC 2023

作者： Rajeshkumar, M. Arun Raj, L. B.S. Abdur Rahman Crescent Institute of Science and Technology Department of Computer Science and Engineering Chennai India

ISBN: (数字)9798350305449

ISBN: (纸本)9798350305449

Object Detection technique is extremely related among the computer Vision field. Object detection empowers detecting case in the perception of various objects from videos plus imagery as well as videos or pre-recordings video. This will find out the variety of individuality of imagery more willingly than the recognition of mechanism of objects and will provides a logical as well as perceptive of images in efficient manner deeply related to human vision works. Object detection has recognized with significant notice from researchers for the motive of its very close relationship with image detection. Machine learning algorithms interrelated to Convolutional Neural Networks (CNNs) has been currently investigated in a several times of real time object detection applications through the source of various articles. In this study, we have analyzed about various algorithm implemented in the concept of real-time object detection as well as importance and drawbacks of specific algorithm of the existing works in the form of a comparative table. The results of this investigation indicate that the machine learning algorithms are very successful in detecting the real time object. © 2023 IEEE.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Enhanced Human Action Recognition with Ensembled DTW Loss Function in CNN LsTM Architecture 3rd

Enhanced Human Action Recognition with Ensembled DTW Loss Fu...

引用

3rd International Conference on sustainable Expert systems, ICsEs 2022

作者： Ram, D. Dinesh Muthukumaran, U. Fatima, N. sabiyath Department of Computer Science and Engineering B.S. Abdur Rahman Crescent Institute of Science and Technology Chennai India

ISBN: (纸本)9789811978739

Human action recognition is a concept that involves acquiring information b.s.d on the sequence of movements by the target. This recognition model is used to recognise elderly people's actions to monitor their anomalies and provide appropriate guidance as soon as possible to avoid further disaster. This algorithm involves refining HAR and provides precise decisions to predict any emergencies while handling elderly people. This algorithm uses Dynamic Time Warping as a loss function ensembled with mean absolute error in CNN LsTM neural network. DTW is used in the loss function to get a relationship between two sequences of actions considered as waves. This results in performing better at any speed of action as long as the sequence of movements for action remains the same. The variations in the change of the sequence of movements for action are further countered by mean absolute error to perform better in any variation of an action. Thus, it provides an accuracy of 75% for the taken dataset which is way more than the accuracy produced by DTW or MAE as a separate loss function. © 2023, The Author(s), under exclusive license to springer Nature singapore Pte Ltd.

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

Advanced Vehicle Detection Heads-Up Display with TensorFlow Lite 3rd

Advanced Vehicle Detection Heads-Up Display with TensorFlow ...

引用

3rd International Conference on sustainable Expert systems, ICsEs 2022

作者： Haris, K. Mohamed Fatima, N. sabiyath Albeez, syed Abdallah Department of Computer Science and Engineering B.S. Abdur Rahman Crescent Institute of Science and Technology Chennai India

ISBN: (纸本)9789811978739

The real-time heads-up display (HUD) will enable the human operator to have a new sense and perspective to driving. The proposed Heads-up display (HUD) detects vehicles, civilians, and obstacles by allowing the user to stay focused on driving. It has a navigator map to keep the user aware of their current location and helps to track their movement. It also tracks speed with an overlayed speedometer and offers fail-safe emergency protocol to alert the emergency contacts when a crash is detected. All these features are accessible by almost everyone who have access to a mobile device and is very cost-effective. The Heads-up display serves as an effective medium between the car and human operator. The operator can set any of the three models: ssD, YOLO v2 and v3 b.s.d on their device performance level. If it is a lower-end device, the ssD model performs faster and efficient. If it is a high-end device, the performance can be improved by using YOLO higher version models. © 2023, The Author(s), under exclusive license to springer Nature singapore Pte Ltd.

关键词： Object detection

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：