检索结果-内蒙古大学图书馆

VisioSLR: A Vision Data-Driven Framework for Sign Language Video Recognition and Performance Evaluation on Fine-Tuned YOLO Models

引用

Procedia computer Science 2025年 257卷 85-92页

作者： Leila Ismail Nada Shahin Henok Tesfaye Alain Hennebelle Intelligent Distributed Computing and Systems (INDUCE) Lab Department of Computer Science and Software Engineering College of Information Technology United Arab Emirates University United Arab Emirates Independent Researcher

With the emergence of AI for good, there has been an increasing interest in building computer vision data-driven deep learning inclusive AI solutions. Sign language Recognition (SLR) has gained attention recently. It is an essential component of a sign-to-text translation system to support the deaf and hard-of-hearing population. This paper presents a computer VISIOn data-driven deep learning framework for Sign Language video Recognition (VisoSLR). VisioSLR provides a precise measurement of translating signs for developing an end-to-end computational translation system. Considering the scarcity of sign language datasets, which hinders the development of an accurate recognition model, we evaluate the performance of our framework by fine-tuning the very well-known YOLO models, which are built from a signs-unrelated collection of images and videos, using a small-sized sign language dataset. Gathering a sign language dataset for signs training would involve an enormous amount of time to collect and annotate videos in different environmental setups and multiple signers, in addition to the training time of a model. Numerical evaluations of VisioSLR show that our framework recognizes signs with a mean average precision of 97.4%, 97.1%, and 95.5% and 11, 12, and 12 milliseconds of recognition time on YOLOv8m, YOLOv9m, and YOLOv11m, respectively.

关键词： Artificial intelligence computer vision Deep learning Fine-tuning Machine translation Neural machine translation Sign language Sign language recognition Transfer learning YOLO

来源：评论

学校读者我要写书评

暂无评论

Refractive Index Biosensor–Based Detection of Mycobacterium Tuberculosis Using Sea Lion Political Optimizer and Deep Learning

引用

Plasmonics 2025年 1-19页

作者： Ragavan, V. K. Senthil Nithya, N.S. Vengala, Anantharamaiah Subramanian, Balambigai Ambhika, C. Sri Shanmugha College of Engineering and Technology Tamil Nadu Sankari Salem (D.T.) 637 304 India Department of Software Systems School of Computer Science and Engineering Vellore Institute of Technology Tamil Nadu Vellore India Department of C S E K L E F Andhra Pradesh Vaddeswaram Guntur 522502 India Dept of ECE Karpagam College of Engineering Tamil Nadu Coimbatore India Department of AIML R.M.D Engineering College RSM Nagar Tamil Nadu Kavarapetai India

Mycobacterium tuberculosis (MTB) detection is a major global health concern, requiring accurate and reliable diagnostic tools. According to the World Health Organization (WHO), the global burden of TB is staggering, with approximately 10 million new cases reported annually, leading to a devastating 1.5 million fatalities worldwide every year. The timely detection and treatment of TB are critical components in the fight against this disease, as they play a vital role in preventing its transmission, minimizing the risk of complications and deaths, and enhancing the overall quality of life for affected individuals. Hence, this study presents a novel Sea Lion Political Optimizer_Deep Maxout Network (SLnPO_DMN) for detecting MTB, leveraging a robust processing pipeline to achieve exceptional accuracy. The proposed pipeline involves a series of meticulous steps, including data pre-processing by Min–Max normalization to scale the data to a common range. To enhance model generalizability, data augmentation is carried out by Bootstrap, generating additional training samples by resampling the original data with replacement. The augmented data is then fed into a DMN, optimized by the SLnPO, a hybrid algorithm combining the strengths of sea lion optimization (SLnO) and political optimizer (PO). The SLnPO_DMN model is trained to identify and learn intricate patterns within the data, ultimately attaining optimal performance and demonstrating exceptional capability in detecting Mycobacterium tuberculosis with high accuracy and reliability. The SLnPO_DMN is evaluated using a comprehensive set of metrics, including accuracy, sensitivity, and specificity. The results demonstrate the effectiveness of the SLnPO_DMN model achieving an accuracy (99.12%), sensitivity of (98.74%), and specificity of (98%). © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2025.

关键词： Deep maxout network Min–Max normalization Mycobacterium tuberculosis Political optimizer Sea lion optimization

来源：评论

学校读者我要写书评

暂无评论

Harnessing the Power of AI with Deep Learning: Innovations and Implications

Harnessing the Power of AI with Deep Learning: Innovations a...

引用

2025 IEEE International Conference on Computational, Communication and Information Technology, ICCCIT 2025

作者： Upender, Talluri Sathiyamoorthy, C. Udayakumar, K. Kirubakaran, D. Purushothaman, V. Raja, Kiruba Thangam CMR College of Engineering and Technology Department of CSE Medchal Telangana Hyderabad India Department of Commerce Tamil Nadu Chennai India Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology Department of CSE Tamil Nadu Chennai India R.M.K. Engineering College Department of Computer Science and Engineering Tamil Nadu Chennai India Chennai Institute of Technology Department of Electronics and Communication Engineering Tamil Nadu Chennai India VIT University Department of Software and Systems Engineering Tamil Nadu Vellore India

ISBN: (纸本)9798331512965

The rapid advancements in artificial intelligence (AI) and deep learning have revolutionized various sectors, enabling unprecedented levels of innovation and efficiency. This paper delves into the transformative impact of deep learning across multiple domains, highlighting key innovations and their broader implications. By examining cutting-edge applications in healthcare, finance, autonomous systems, and natural language processing, the study elucidates how deep learning models, such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs), are pushing the boundaries of what's possible. Moreover, the paper explores the integration of deep learning with emerging technologies such as the Internet of Things (IoT), edge computing, and quantum computing, underscoring the synergistic potential of these combinations. The role of deep learning in predictive analytics and personalized solutions is discussed, showcasing its ability to enhance decision-making processes and improve user experiences. Furthermore, the paper discusses the ethical considerations, challenges, and future prospects associated with the widespread adoption of deep learning technologies. The insights provided aim to foster a deeper understanding of how harnessing AI with deep learning can drive progress while addressing critical societal and technological issues. © 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

A Comparison of Deep Learning and Machine Learning Methods for Water Quality Prediction

A Comparison of Deep Learning and Machine Learning Methods f...

引用

Intelligent Data Communication Technologies and Internet of Things (IDCIoT), International Conference on

作者： N. Srinivasa Gupta B. Valarmathi Ishaan Hiwanj Department of Manufacturing Engineering School of Mechanical Engineering Vellore Institute of Technology Vellore Tamil Nadu India Department of Software and Systems Engineering School of Computer Science Engineering and Information Systems Vellore Institute of Technology Vellore Tamil Nadu India Department of Information Technology School of Computer Science Engineering and Information Systems Vellore Institute of Technology Vellore Tamil Nadu India

ISBN: (数字)9798331527549

ISBN: (纸本)9798331527556

Predicting water quality is essential to preserving human health and environmental sustain ability. Traditional water quality assessment methods often face scalability and real-time monitoring limitations. With accuracies of 62%, 72 %, 83 %, 69%, 63 %, 66%, 71 %, 63 %, and 64%, respectively, the current techniques utilized were Logistic Regression, Decision Trees, Random Forest Regressor, Extreme Gradient Boosting, Naive Bayes, K-nearest neighbors, Support Vector Machine, AdaBoost, and Bagging [9]. This study addresses these challenges by leveraging Adaptive Synthetic Sampling (ADASYN) to balance the dataset and evaluating model performance on datasets of 5,000 and 10,000 entries per class. A robust dataset obtained from Kaggle was used, with five models - Long Short-Term Memory (LSTM), Feed Forward Neural Network (FFNN), Categorical Boosting (CatBoost), Extreme Gradient Boosting (XGBoost), and Random Forest - evaluated and compared. The proposed methods demonstrate significant improvements in accuracy, with XGBoost achieving the highest accuracy of 95.53%, followed by Random Forest at 93.98%. This work underscores the importance of advanced machine learning techniques in addressing the limitations of traditional methods, enhancing accuracy, scalability, and adaptability in water quality prediction. These findings contribute to advancing environmental monitoring and management practices with reliable, data-driven insights.

关键词： Support vector machines Adaptation models Accuracy Scalability Water quality Boosting Reliability Random forests Long short term memory Regression tree analysis

来源：评论

学校读者我要写书评

暂无评论

Optimizing Existing SHM systems: Retasking as a Self-Healing Solution for Improved Fault Tolerance

Optimizing Existing SHM Systems: Retasking as a Self-Healing...

引用

Emerging Smart Computing and Informatics (ESCI), Conference on

作者： Aaron Mano Cherian D. Ajitha Avanish Gouraha Dipanshu Mandal School of Computer Science and Engineering (SCOPE) Vellore Institute of Technology Vellore Tamilnadu India Department of Software Systems School of Computer Science and Engineering (SCOPE) Vellore Institute of Technology Vellore Tamilnadu India

ISBN: (数字)9798331515683

ISBN: (纸本)9798331515690

This paper focuses on self-healing algorithms in structural health monitoring (SHM) systems centered around the enhancement of resilience and adaptability of the systems. In this study, imports from existing methods (clustering, Fault Tolerant Multiple Redundancy (FTMR) and reinforcement learning) are analyzed against the choice of creating a novel retasking algorithm designed for dynamic resource redistribution and optimal monitoring coverage. Unlike conventional methods, retasking will allow adapting the coverage in real time, whereby system down time will be reduced, with less computational load achieved through task redistribution through functional sensors. Findings showed that retasking improved reliability and scalability of the SHM systems drastically, providing a simple yet powerful resolution towards modern infrastructure monitoring. This study stresses the retasking capability to redefine self-healing in the SHM systems for future directions in infrastructure safety.

关键词： Training Heuristic algorithms Fault tolerant systems Redundancy Clustering algorithms Reinforcement learning Solids Real-time systems Monitoring Stress

来源：评论

学校读者我要写书评

暂无评论

Real-Time Freeway Traffic Anomalous Event Detection System via Radar Detector Sensors 25

Real-Time Freeway Traffic Anomalous Event Detection System v...

引用

Proceedings of the ACM/IEEE 16th International Conference on Cyber-Physical systems (with CPS-IoT Week 2025)

作者： Austin Coursey Junyi Ji Zhiyao Zhang William Barbour Marcos Quinones-Grueiro Tyler Derr Gautam Biswas Daniel B. Work Institute for Software Integrated Systems Vanderbilt University Nashville TN USA and Department of Computer Science Vanderbilt University Nashville TN USA Institute for Software Integrated Systems Vanderbilt University Nashville TN USA and Department of Civil and Environmental Engineering Vanderbilt University Nashville TN USA Institute for Software Integrated Systems Vanderbilt University Nashville TN USA Department of Computer Science Vanderbilt University Nashville TN USA Institute for Software Integrated Systems Vanderbilt University Nashville TN USA Department of Computer Science Vanderbilt University Nashville TN USA and Department of Civil and Environmental Engineering Vanderbilt University Nashville TN USA

ISBN: (纸本)9798400714986

Undetected anomalous incidents on the freeway, such as crashes, can cause increased congestion and secondary crashes. To address this problem, we recently released a real freeway traffic anomalous event detection dataset. However, our evaluation was limited to a static week of traffic data. This demo presents a live anomaly detection dashboard, processing sensor data with a 1-minute delay. It has interactive components to adjust the anomaly detection performance, providing a realistic evaluation of graph anomaly detection methods for long-term analysis and method improvements.

关键词： anomaly detection

来源：评论

学校读者我要写书评

暂无评论

Exploratory Data Analysis and Predicting Indian Crime Against Women Using Machine Learning Approaches

Exploratory Data Analysis and Predicting Indian Crime Agains...

引用

Interdisciplinary Approaches in Technology and Management for Social Innovation (IATMSI), IEEE Conference on

作者： N. Srinivasa Gupta B. Valarmathi K. Santhi T.Chellatamilan Sameeksha Nanda Department of Manufacturing School of Mechanical Engineering Vellore Institute of Technology Vellore Tamil Nadu India Department of Software and Systems Engineering School of Computer Science Engineering and Information Systems Vellore Institute of Technology Vellore Tamil Nadu India Department of Analytics School of Computer Science and Engineering Vellore Institute of Technology Vellore Tamil Nadu India Department of Information Technology School of Computer Science Engineering and Information Systems Vellore Institute of Technology Vellore Tamil Nadu India

ISBN: (数字)9798331521691

ISBN: (纸本)9798331521707

Crime forecasting is a critical endeavor aimed at preventing future criminal actions, specifically the rising rate of violence against women in India. This study proposes the model to predict crimes committed against women using data from the National Crime Records Bureau's (NCRB) monthly crime report. The dataset encompasses district and state-wise crime activities from 2001 to 2014. The proposed work objectives are identifying factors influencing these crimes and developing predictive models using machine learning (ML). The study employs Exploratory Data Analysis (EDA), to enhance the predictive power of the models. The suggested study uses XGBoost, Random Forest (RF), and Neural Network (NN). The neural network algorithm outperforms the Random Forest and XGBoost techniques among these three. The Random Forest algorithm has the worst performance. After comparing the accuracies of various methods, such as XGBoost (92%), and Random Forest (89.84%) classifiers using imbalanced datasets (without employing Synthetic Minority Oversampling Technique (SMOTE) technique), it was found that the neural network algorithm (95.81%) delivers the highest accuracy. This study looks at model accuracies and finds that, after applying the SMOTE algorithm, neural network algorithms performed best, with an accuracy of 96.76%, outperforming the accuracies of the remaining models, which included the Random Forest (93.08%) and XG-Boost (95%).

关键词： Radio frequency Technological innovation Accuracy Machine learning algorithms Data analysis Predictive models Prediction algorithms Data models Classification algorithms Random forests

来源：评论

学校读者我要写书评

暂无评论

Enhancing Human Posture Detection with Hybrid Deep VGG16 and Attention Mechanism Network

引用

Procedia computer Science 2025年 258卷 1209-1218页

作者： Ogundokun Roseline Oluwaseun Rytis Maskeliūnas Federick Oscar Oluwakayode A. Oki Department of Software Engineering Kaunas University of Technology Kaunas 44249 Lithuania Department of Computer Science Landmark University Omu-Aran Nigeria Department of Computer Systems Engineering Tshwane University of Technology (TUT) Pretoria South Africa Adioo Technology Abuja Nigeria Department of Information Technology East London Walter Sisulu University South Africa

Human Posture Detection (HPD) and classification are the initial steps toward various computer vision applications related to security, advertisement, and healthcare. Even though more than three decades have passed, most methods have been focused only on detecting standing people. However, in real applications, human postures may be significantly different, such as standing, sitting, lying, and crouching, and the shapes of a human are varied with viewpoints, thus making detection and classification difficult. Moreover, the gradual transition from one posture to another further complicates the determination of the number of postures to be classified. This paper uses a pre-trained model to determine the number of unique human postures in activities. We approached PC as a multiclass detection problem using the very deep VGG 16 Convolutional Neural Network (CNN) empowered by an attention mechanism. Extract frames from video data recorded, followed by the derivation of features through a convolutional neural network. Classify different postures—bending, exercise, lying, sitting, and standing—with VGG 16 and attention mechanism. It is trained on the ETRI-Activity3D-LivingLab Dataset. Our proposed (VEACT-CNN) framework has shown stability in HPD with an overall accuracy of 95% and an F-measure of 98% while generating a very low level of false alarms rate (FAR) of 0.02 (2%). For comparison purposes, the accuracy obtained by other models on the same dataset is VGG 16.

关键词： Vision-based human detection Posture classification VGG 16 Deep learning Pre-trained model Human posture detection

来源：评论

学校读者我要写书评

暂无评论

Gender Classification from Pashto Handwritten Text Images

引用

IEEE Access 2025年 13卷 81736-81754页

作者： Sultan, Khan Ahmad, Riaz Muhammad, Siraj Almutairi, Sulaiman Abohashrh, Mohammed Ullah, Khalil Namoun, Abdallah Hussain, Ibrar Shaheed Benazir Bhutto University Department of Computer Science Dir Upper Khyber Pakhtunkhwa Sheringal75660 Pakistan Qassim University College of Applied Medical Sciences Department of Health Informatics Qassim52571 Saudi Arabia King Khalid University College of Applied Medical Sciences Department of Basic Medical Sciences Abha61421 Saudi Arabia University of Malakand Department of Software Engineering Dir Lower Khyber Pakhtunkhwa Malakand18800 Pakistan Islamic University of Madinah AI Center Faculty of Computer and Information Systems Madinah42351 Saudi Arabia Shaheed Benazir Bhutto University Sheringal Quality Enhancement Cell Khyber Pakhtunkhwa Dir Upper75660 Pakistan

computer vision (CV) is a subfield of computer science that enables machines to perceive, interpret, and understand visual data. It combines image processing, analysis, and machine learning to extract meaningful insights from images and videos. One key application of CV is the classification of gender from handwritten text images-a complex task, especially in languages such as Pashto due to the intricacy of the script and the limited resources. This research presents a novel approach to gender classification using Pashto Handwritten Text Images (PHTI). The study evaluates and compares the performance of traditional machine learning and modern deep learning models: Support Vector Machine (SVM) and You Only Look Once version 8 (YOLO8), a CNN-based object detection framework. The PHTI dataset, introduced in this study, consists of 36,086 text line images written by 200 male and 200 female native Pashto writers. The goal is to assess how effectively SVM and YOLO8 can classify gender based on visual handwriting patterns in the Pashto language. The experimental results reveal that YOLO8 outperforms SVM in classification accuracy. Although SVM, a widely used machine learning model, achieved an accuracy of 61.29%, YOLO8 reached a significantly higher accuracy of 77% in the same data set. This performance gap highlights the advantages of using advanced deep learning models for complex pattern recognition tasks in CV. This research provides a pioneering contribution by focusing on Pashto, a low-resource language, in the context of gender classification using handwritten text images. It opens new pathways for applying CV in regional languages and supports the development of intelligent systems capable of handling various scripts. © 2013 IEEE.

关键词： Machine vision

来源：评论

学校读者我要写书评

暂无评论

LED there be DoS: Exploiting variable bitrate IP cameras for network DoS

arXiv

引用

arXiv 2025年

作者： Goldberg, Emmanuel Brodt, Oleg Elyashar, Aviad Puzis, Rami Department of Software and Information Systems Engineering Ben-Gurion University of the Negev Israel Cyber Labs@BGU Ben-Gurion University of the Negev Israel Department of Computer Science Shamoon College of Engineering Israel

Variable-bitrate video streaming is ubiquitous in video surveillance and CCTV, enabling high-quality video streaming while conserving network bandwidth. However, as the name suggests, variable-bitrate IP cameras can generate sharp traffic spikes depending on the dynamics of the visual input. In this paper, we show that the effectiveness of video compression can be reduced by up to 6× using a simple laser LED pointing at a variable-bitrate IP camera, forcing the camera to generate excessive network traffic. Experiments with IP cameras connected to wired and wireless networks indicate that a laser attack on a single camera can cause significant packet loss in systems sharing the network with the camera and reduce the available bandwidth of a shared network link by 90%. This attack represents a new class of cyber-physical attacks that manipulate variable bitrate devices through changes in the physical environment without a digital presence on the device or on the network. We also analyze the broader view of multidimensional cyberattacks that involve both the physical and digital realms and present a taxonomy that categorizes attacks based on their direction of influence (physical-to-digital or digital-to-physical) and their method of operation (environment-driven or device-driven), highlighting multiple areas for future research. © 2025, CC BY-NC-ND.

关键词： Video streaming

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：