检索结果-内蒙古大学图书馆

2022 International Conference for Advancement in Technology, ICONAT 2022

作者： Hasan, Nazmul Nene, Manisha J. Defence Institute of Advanced Technology Computer Science and Engineering Department Pune India

ISBN: (纸本)9781665425773

Children with Autism Spectrum Disorder (ASD) suffer most in communication and behavioral aspect which hampers their regular basic education and cognitive development. Technological development enhanced a lot in every sector where educational sector is no exception of that. The cognitive development of the children with ASD can be taken care of effectively by using of advanced information technology. This study has focused on the prospect of designing the digital learning platform for the children with ASD to fulfill the existing research gap for developing persuasive learning mechanism for the children with ASD. This study has proposed an agent-based model for developing a persuasive learning tool where Sanket is an agent of a touch screen based application which pursues the children with ASD regarding the learning of basic education (numbers, alphabets), which forms a basis to enable Artificial Intelligence for automation of the learning process and help the children with ASD. © 2022 IEEE.

关键词： Touch screens

来源：评论

学校读者我要写书评

暂无评论

DropAttack: A Random Dropped Weight Attack Adversarial Training for Natural Language Understanding

引用

IEEE/ACM Transactions on Audio Speech and Language Processing 2024年 32卷 364-373页

作者： Ni, Shiwen Li, Jiawen Yang, Min Kao, Hung-Yu Chinese Academy of Sciences Shenzhen Institute of Advanced Technology Shenzhen518000 China Guangdong Ocean University Maritime College Zhanjiang524000 China National Cheng Kung University Department of Computer Science and Information Engineering Tainan701401 Taiwan

Adversarial training has been proven to be a powerful regularization technique to improve language models. In this work, we propose a novel random dropped weight attack adversarial training method (DropAttack) for natural language understanding. Our DropAttack improves the generalization of models by minimizing the internal adversarial risk caused by a multitude of attack combinations. Specifically, DropAttack enhances the adversarial attack space by intentionally adding worst-case adversarial perturbations to the weight parameters and randomly dropping the specific proportion of attack perturbations. To extensively validate the effectiveness of DropAttack, 12 public English natural language understanding datasets were used. Experiments on the GLUE benchmark show that when DropAttack is applied only to the finetuning stage, it is able to improve the overall test scores of the BERT-base pre-trained model from 78.3 to 79.7 and RoBERTa-large pre-trained model from 88.1 to 88.8. Further, DropAttack also significantly improves models trained from scratch. Theoretical analysis reveals that DropAttack performs potential gradient regularization on the input and weight parameters of the model. Moreover, visualization experiments show that DropAttack can push the minimum risk of the neural network to a lower and flatter loss landscape. © 2014 IEEE.

关键词： Speech processing

来源：评论

学校读者我要写书评

暂无评论

Forensic Analysis of Windows 11 Prefetch Artifact 4

Forensic Analysis of Windows 11 Prefetch Artifact

引用

4th IEEE Bombay Section Signature Conference, IBSSC 2022

作者： Budhrani, Akash Singh, Upasna Singh, Bhupendra Defense Institute of Advanced Technology Computer Science and Engineering Department Pune India

ISBN: (纸本)9781665492911

The Operating System creates numerous objects to improve its efficiency and user experience and such objects are called artifacts. These artifacts record crucial data about the user activity. Such artifacts are the start point of any investigation as they can be an additional lead to a forensic triage. Prefetch file is one among various objects, presence of which confirms the execution of a particular application. Prefetch gives additional inside for the purpose of investigation. Thus, this paper brings out the forensic value of it, the tools required to decode the information it contains and also look in various caveats in interpreting this artifact to learn its strength and weaknesses to properly incorporate in support of opinion derived by the analyst. In this work, Prefetch is forensically examined to bring out its forensic value, knowledge it contains and all of that in whole or in parts can be used to help advance in investigation. Paper also brings out the difference in format of this artifact among various version of Windows OS. © 2022 IEEE.

关键词： Windows operating system

来源：评论

学校读者我要写书评

暂无评论

Enhancing Lung Cancer Detection Accuracy through the Integration of Ant Colony Optimization and Neural Network

Enhancing Lung Cancer Detection Accuracy through the Integra...

引用

2024 International Conference on Emerging Innovations and advanced Computing, INNOCOMP 2024

作者： Bhati, Sunita Singh, Jagendra Shelke, Nitin Arvind Nagda, Vinita Sahu, Hemant Bhatt, Monica JECRC University Jaipur India School of Computer Science Engineering & Technology Bennett University Greater Noida India Mohan Lal Sukhadia University Department of Computer Science Udaipur India Department of CS & IT Udaipur India Geetanjali Institute of Technical Studies Udaipur India

ISBN: (纸本)9798350376470

The combination of machine learning algorithms and optimization techniques shows great potential in lung cancer detection. The current study explores the use of Convolutional Neural Networks (CNN) and VGG16 models with Ant Colony Optimization (ACO) for better performance of predictive modeling in lung cancer detection. In the study, a dataset of 4,236 images from Computed Tomography scans was used. The dataset comprised both CT scans of cancerous lungs at different stages and healthy, non-cancerous lung images. First, ACO was used to select important features, then, the selected features were used to train the two models. As a result, high accuracy of the models was observed, with CNN having 95.46% accuracy and VGG16 - 97.87%. Comprehensive evaluation scores, including precision, recall, and F1-score, also showed the models' high-performance level in differential analysis of disease and healthy lung tissue. The results further emphasize that the use of advanced machine learning algorithms with optimization techniques has the potential to revolutionize lung cancer detection, providing the medical community with powerful tools both for early detection and personalized treatment. However, continued efforts in research and development are needed to refine the performance of predictive models and ultimately better fight lung cancer and improve prognosis for patients. © 2024 IEEE.

关键词： Lung cancer

来源：评论

学校读者我要写书评

暂无评论

Rectifying Irregular Boundaries Stereoscopic Images Using Deep Learning and Evolutionary Algorithms 22

Rectifying Irregular Boundaries Stereoscopic Images Using De...

引用

22nd IEEE International Multi-Conference on Systems, Signals and Devices, SSD 2025

作者： Wahsh, Muntasser A. Hussain, Zainab M. Informatics Institute for Postgraduate Studies (IIPS) Department of Computer Science Baghdad Iraq Al-Iraqia University Department of Political Science Baghdad Iraq

ISBN: (纸本)9798331542726

Application of stereoscopic technology has increased significantly in several industries, including the media and entertainment sectors, medical imaging, and virtual reality. However, the overall viewing experience and the feeling of depth can be negatively impacted by uneven boundaries in stereoscopic images. To have a realistic and immersive stereoscopic experience, fixing these wavy edges is an essential process. This work present a distinctive approach for optimizing the rectangular bounds of stereoscopic images using a hybrid of deep learning and evolutionary algorithms. The suggested method uses a genetic algorithm to optimize the parameters of a convolutional neural network (CNN) to extract useful features for residual regression. Due to the lack of compatible datasets, a set of stereoscopic images was constructed from monocular images by calculating the depth of the objects using MiDaS pretrained depth estimation algorithm. Several metrics, including as mean squared error, peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), and Fréchet inception distance (FID), were used to evaluate the performance of the suggested technique. The experimental findings show that the proposed method effectively optimizes rectangular borders with excellent accuracy, even for images with complicated shape and orientations. Various parameter settings were also tested, revealing that increasing the number of generations and population size contributes to improved performance. The combination of the deep learning algorithm, specifically the CNN, and the optimization algorithm, specifically the genetic algorithm (GA), leads to an increase in the accuracy of rectified images. Notably, the average PSNR reached 24.42, SSIM measured at 0.7793, and FID scored 18.26. The proposed method holds promise for a wide range of applications, including image cropping, object detection, and image segmentation, among others. By effectively optimizing rectangular boundaries in ster

关键词： Deep learning Depth estimation Genetic algorithm Image rectifying Monocular to stereoscopic Stereoscopic images

来源：评论

学校读者我要写书评

暂无评论

Real-Time Eyes Monitoring System Using OpenCV Autonomous filtering using RNN 3

Real-Time Eyes Monitoring System Using OpenCV Autonomous fil...

引用

3rd International Conference for Advancement in Technology, ICONAT 2024

作者： Prajapati, Dhruvansh Joshi, Jaynil Das, Nibedita Satapathy, Santosh Pandit Deendayal Energy University Information & Communication Technology Gujarat Gandhinagar India Institute of Advanced Research Department of Computer Science and Engineering Gujarat Gandhinagar India

ISBN: (纸本)9798350354171

This paper introduces a novel Real-Time Eye Gazing and Monitoring System aimed at objectively and continuously assessing individuals across various environmental conditions. Addressing the challenge of accurate performance in diverse lighting conditions, the system employs Recurrent Neural Network (RNN) trained filters to enhance eye detection accuracy. Traditional methods often falter in non-ideal lighting, prompting the need for adaptive solutions. Through extensive experimentation, optimal filter combinations were identified, ensuring robustness across a spectrum of scenarios. Implementation via an OpenCV facilitates efficient processing without compromising detection accuracy. This paper achieved the list of main filters and their combinations which can be helpful to detect eyes in different environmental conditions. © 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

Photovoltaic Module Parameter Estimation by Using Wild Horse Optimizer 3

Photovoltaic Module Parameter Estimation by Using Wild Horse...

引用

3rd International Conference on Machine Learning and Data Engineering, ICMLDE 2024

作者： Kumar, Anuj Sharma, Tina Singh, Manoj K. Pant, Sangeeta Sharma, Hitesh Kumar Pune411044 India Department of Mathematics Birla Institute of Technology and Science Pilani India School of Computer Science Engineering and Technology Bennett University Greater Noida201310 India Pune412115 India School of Computer Science University of Petroleum & Energy Studies Dehradun248007 India

The development of highly accurate models is essential for optimizing the performance and efficiency of photovoltaic (PV) modules, which are integral to renewable energy systems. In this context, metaheuristic algorithms have become indispensable tools for tackling complex engineering challenges, particularly in the areas of evaluation, simulation, and parameter estimation. These algorithms excel at finding optimal or near-optimal solutions for problems where traditional methods may struggle due to their complexity or computational intensity. This article presents a novel framework for estimating the parameters of PV modules using the Wild Horse Optimizer (WHO), a recently introduced metaheuristic algorithm known for its robustness and efficiency. Specifically, the framework is applied to the Double Diode Model (DDM) of a solar cell, a widely recognized model for representing the electrical behavior of PV modules. The simulation results clearly illustrate that the proposed framework based on WHO algorithm offers significant improvements in the accuracy and effectiveness of parameter estimation for PV modules, compared to Artificial Bee Swarm Optimization (ABSO), the Flower Pollination Algorithm (FPA), and several others, thereby validating the proposed approach. © 2024 The Authors. Published by ELSEVIER B.V.

关键词： Direction of arrival

来源：评论

学校读者我要写书评

暂无评论

Abstractive Text Summarization Using Attention-based Stacked LSTM 5

Abstractive Text Summarization Using Attention-based Stacked...

引用

5th International Conference on Computational Intelligence and Communication Technologies, CCICT 2022

作者： Singh, Mimansha Yadav, Vrinda Centre for Advanced Studies Department of Computer Science and Engineering Uttar Pradesh Lucknow226031 India

ISBN: (数字)9781665472241

ISBN: (纸本)9781665472241

Every day, the amount of textual data created increases exponentially, both in terms of complexity and volume. Massive amounts of information are generated by social media, news articles, emails, text messages and other resources, making it difficult to read lengthy language materials. Our main objective in the paper is to obtain a short understandable and fluent abstractive summary of any given text. The Abstractive Text Summarizer automatically gives the summary of the text by generating new phrase, rephrasing or using the new words which are not present in the original text. In this paper, a machine learning architecture i.e. Stacked LSTM based on attention mechanism using Sequence-to-Sequence model is proposed, to generate the summary using abstractive approach for Amazon reviews of fine foods dataset. Our approach allows the model to accept content and provide a concise summary that may clearly describe the gist of the original text. The experiments on Amazon reviews of fine foods dataset show that our model obtained BLEU Score as 0.91 for a test set. © 2022 IEEE.

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

Drone/UAVs Surveillance feeds Video Object Recognition Using Transfer Learning 1

Drone/UAVs Surveillance feeds Video Object Recognition Using...

引用

1st International Conference on Futuristic Technologies, INCOFT 2022

作者： Kumar, Sushil Kumar, C.R.S. Defense Institute of Advanced Technology Computer Science and Engineering Department Pune India

ISBN: (纸本)9781665450461

The objective of this research is to demonstrate the use of a convolutional neural network (CNN) for object detection (OD) on drone videos (CNN). The goal of the study is to determine how well the YOLO V3 (Y-V3) and feature extractor recognise objects in drone-shot video frames the *** to the study, application of several target identification algorithms to 'regular' images captured by 'normal' cameras has different efficiency impacts on the quantity, precision, and target performance consumption, as well as when applying the technique to image data collected by drones. Unmanned aerial vehicles (UAVs), a very active domain in this field, are crucial to the implementation of any robot's complete autonomy due to its enhanced capability of OD. We have conducted numerous experiments to address our functional issues to investigate the efficacy of the most sophisticated target detection algorithm in the image data acquired by UAV. In this study, the algorithm's performance metrics are tallied after being tested on 1099 photos. © 2022 IEEE.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Forensic Techniques to Detect Hidden Data in Alternate Data Streams in NTFS 4

Forensic Techniques to Detect Hidden Data in Alternate Data ...

引用

4th IEEE Bombay Section Signature Conference, IBSSC 2022

作者： Hermon, Rahul Singh, Upasna Singh, Bhupendra Defense Institute of Advanced Technology Computer Science and Engineering Department Pune India

ISBN: (纸本)9781665492911

Alternate Data Streams (ADS) have been a feature of the New Technology File System (NTFS) since its introduction in 1993. Alternate Data Streams (ADS) were introduced to address compatibility within the existing Operating Systems. Lately Hackers/Cyber Criminals have used Alternate Data Streams (ADS) as a means for launching Cyber- Attacks. Alternate Data Streams (ADS) allow data hiding, same being difficult to detect without adequate knowledge. In this paper we shall bring out the various Forensic techniques in which hidden data in Alternate Data Streams (ADS) can be detected. Finally, we compared the Forensic techniques to detect data hidden in Alternate Data Streams (ADS) in both Windows 10 and 11 Operating System. © 2022 IEEE.

关键词： Network security

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：