检索结果-内蒙古大学图书馆

2022 International Conference on Artificial Intelligence and data Engineering, AIDE 2022

作者： Bijur, Gururaj Ramakrishna, M. Karunakar, A.K. Manipal Institute of Technology Department of Computer Science & Engineering MAHE Manipal Karnataka Udupi India Manipal Institute of Technology Department of Data Science & Computer Applications MAHE Manipal Karnataka Udupi India

ISBN: (纸本)9781665493048

HEVC is the new coding standard aiming to provide better coding efficiency compared to its previous standard. Meantime this increases the computational complexity. In HEVC, there are 35 intra prediction modes, where as in H.264/A VC there were only 9 intra prediction modes. In this paper, we proposed a fast rough mode decision (RMD) algorithm to decrease the computational complexity using up and left reference pixels. The three-stage search method is implemented in RMD algorithm. The results obtained when compared with HM16.18, shows that the proposed algorithm can save 12.78% to 28.86% of the encoding time with acceptable loss in video quality. © 2022 IEEE.

关键词： Computational complexity

来源：评论

学校读者我要写书评

暂无评论

Fovea Segmentation Using Semi-Supervised Learning 20

Fovea Segmentation Using Semi-Supervised Learning

引用

20th IEEE India Council International Conference, INDICON 2023

作者： Ghosh, Ankita Khose, Sahil Kamath, Yogish S. Kuzhuppilly, Neetha I. R. Harish Kumar, J.R. Manipal Academy of Higher Education Manipal Institute of Technology Dept. of Computer Science and Engineering Manipal India Manipal Academy of Higher Education Manipal Institute of Technology Dept. of Information and Communication Technology Manipal India Manipal Academy of Higher Education Kasturba Medical College Dept. of Ophthalmology Manipal India Manipal Academy of Higher Education Manipal Institute of Technology Dept. of Electrical and Electronics Engineering Manipal India

ISBN: (纸本)9798350305593

Despite the accessibility of retinal fundus images in recent years, fovea segmentation is an exacting task due to the insufficiency of labelled data. In this paper, we propose a deep learning pipeline which utilizes unlabelled data alongside labelled data for the segmentation of fovea. We train 484 labelled images using the Deeplabv3+ architecture and deploy EfficientNet-B3 as the encoder in the framework. Additionally, we introduce semi-supervised learning in our pipeline and train 1200 unlabelled images by generating their pseudo labels. We evaluate our results on the Jaccard index, Dice score, sensitivity, specificity and accuracy. Our Dice score of 82.43% and Jaccard index of 70.52% surpasses the existing methods. We obtain 91.74% sensitivity, 99.75% specificity and 99.57% accuracy. © 2023 IEEE.

关键词： Fovea Segmentation Semi-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Identification of the Video Level for Video Communication in a Resource-Constrained Network

Identification of the Video Level for Video Communication in...

引用

2022 International Conference on Artificial Intelligence and data Engineering, AIDE 2022

ISBN: (纸本)9781665493048

HEVC, which employs layered video coding techniques, is an extension of H.264 Advanced Video Coding. Layered video coding creates a valid bitstream by removing layers. Video streaming via wireless networks is made possible by this capability. The elements that affect the quality of video communication include node mobility, the kind of routing protocols, communication distance, etc. The video adaptation method calculates the number of video levels that will be lost and then extracts at Media Aware Network Elements in accordance with terminal capabilities and resource availability. Building the information needed to select the extraction point or adaption level is the goal of this work. The extraction points for streaming Multi-level Video over the resource constrained network have been calculated in this article. © 2022 IEEE.

关键词： Video streaming

来源：评论

学校读者我要写书评

暂无评论

Prostate Cancer Detection Using Integrated Multi Modal Approaches

Prostate Cancer Detection Using Integrated Multi Modal Appro...

引用

2024 International Conference on Distributed Systems, computer Networks and Cybersecurity, ICDSCNC 2024

作者： Rao, Manjula Gururaj Priyanka, H. Ahamed Shafeeq, B.M. Hemant Kumar Reddy, K. Nmam Institute of Technology Depatment of Computer Science and Engineering Nitte India Pes University Bengaluru India Manipal Institute of Technology Manipal Academy of Higher Education Dept. of Cse Manipal India Vit Ap University Dept. of Cse Inavolu India

ISBN: (纸本)9798350375442

Cancer is one of the terminal diseases that significantly affects an individual's health. Tumours are created when aberrant cells proliferate and divide out of control, posing a threat to neighbouring tissues and organs. Early detection through regular screenings and check-ups can make cancer more treatable. Lots of people throughout are impacted by prostate cancer, the cancer in males that is most frequently diagnosed. The proposed model is a multimodal hybrid approach that utilizes data *** files containing prostate attributes and prostate images. It employs both ML and DL models to classify the data as cancerous or non-cancerous and can further stage the cancerous cells into four stages. The proposed method uses the majority voting system to identify the cancerous or non-cancerous. The model achieves an overall accuracy of 81.48 %, with the ML model at 77.8% and the DL model at 87%. © 2024 IEEE.

关键词： Lung cancer

来源：评论

学校读者我要写书评

暂无评论

Explainable Classification of Macular Degeneration Using Deep Learning 20

Explainable Classification of Macular Degeneration Using Dee...

引用

20th IEEE India Council International Conference, INDICON 2023

作者： Khose, Sahil Ghosh, Ankita Kamath, Yogish S. Kuzhuppilly, Neetha I. R. Kumar, J. R. Harish Manipal Institute of Technology Manipal Academy of Higher Education Dept. of Information and Communication Technology Manipal India Manipal Institute of Technology Manipal Academy of Higher Education Dept. of Computer Science and Engineering Manipal India Kasturba Medical College Manipal Academy of Higher Education Dept. of Ophthalmology Manipal India Manipal Institute of Technology Manipal Academy of Higher Education Dept. of Electrical and Electronics Engineering Manipal India

ISBN: (纸本)9798350305593

Age-related macular degeneration is one of the leading causes of vision loss in individuals. This paper proposes a fundus image-based analysis and automatic grading of AMD using deep learning techniques. We utilize data points from three different datasets and perform augmentation and data sampling to equalise the data distribution. We apply deep learning by proposing EfficientNet-B3 for training a three-class classification model which categorizes the fundus images as having no AMD, mild AMD or severe AMD. The results obtained are evaluated on various metrics and we obtain 93.6% accuracy. We also analyse the trained model by visualising the information learned by it using class activation map algorithms. The accuracy of the proposed method outperforms the majority of the existing methods. The confidence in the proposed model is further increased by the CAM algorithms which indicate that the model predicts the output based on the presence of retinal lesions. © 2023 IEEE.

来源：评论

学校读者我要写书评

暂无评论

Analyzing Effects of Fake Training data on the Performance of Deep Learning Systems

arXiv

引用

arXiv 2023年

作者： Seth, Pratinav Bhandari, Akshat Lakara, Kumud Dept. of Data Science and Computer Applications Dept. of Computer Science and Engineering Manipal Institute of Technology Manipal Academy of Higher Education Manipal India

Deep learning models frequently suffer from various problems such as class imbalance and lack of robustness to distribution shift. It is often difficult to find data suitable for training beyond the available benchmarks. This is especially the case for computer vision models. However, with the advent of Generative Adversarial Networks (GANs), it is now possible to generate high-quality synthetic data. This synthetic data can be used to alleviate some of the challenges faced by deep learning models. In this work we present a detailed analysis of the effect of training computer vision models using different proportions of synthetic data along with real (organic) data. We analyze the effect that various quantities of synthetic data, when mixed with original data, can have on a model's robustness to out-of-distribution data and the general quality of predictions. © 2023, CC BY.

关键词： computer vision

来源：评论

学校读者我要写书评

暂无评论

Secure Online Banking Authentication System Using Time Bound Password

Secure Online Banking Authentication System Using Time Bound...

引用

Computing, Power and Communication Technologies (IC2PCT), IEEE International Conference on

作者： Ronson Danthy K Pratham Pai Vidya Rao Department of Data Science and Computer Applications Manipal Institute of Technology Manipal Academy of Higher Education Manipal India Department of Computer Science and Computer Applications Manipal Institute of Technology Manipal Academy of Higher Education Manipal India

The banking system is a key component of financial transactions and economic expansion in the modern world. The introduction of internet banking, however, has created some brand-new difficulties, particularly in providing safe user authentication. This study investigates the flaws in conventional authentication techniques, including passwords and SMS-based One-Time Passwords (OTP), and suggests a creative fix. By using Time-Based One-Time Password (TOTP) created by an authenticator app and encrypting the user credentials with this as key, the proposed solution improves security. By addressing problems including poor password habits, SMS theft, OTP delivery delays, phishing, and social engineering attacks, this method strengthens the security of online banking. The approach provides a thorough framework for building this secure authentication system by detailing TOTP generation and encryption. This system successfully defends against popular attacks such as replay attacks, brute force attacks, rainbow tables, packet sniffing, and random guessing.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Distance Validation and Role of High Fidelity VR Environments on Embodiment and Enjoyment 3

Distance Validation and Role of High Fidelity VR Environment...

引用

3rd IEEE International Conference on Computational Intelligence and Knowledge Economy, ICCIKE 2023

作者： Kini, Vishwas G. Siddalingaswamy, P.C. Ganeshrao, Shonraj Ballae Manipal Institute of Technology Manipal Academy of Higher Education Dept. of Computer Science and Engg. Manipal India Manipal College of Health Professions Manipal Academy of Higher Education Dept. of Optometry Manipal India UGDX Institute of Technology Hyderabad India

ISBN: (纸本)9798350338263

Gamification, especially in Virtual Reality (VR) have been gaining importance lately for which there arises a need in understanding the interaction mechanism as to which would lead to engagement and motivation. To assess this, our work uses a virtual hand representation to embody the user in three different VR environment of varying interaction fidelity. Through both qualitative and quantitative data gathered, we assess the users' perceived sense of embodiment, engagement, and motivation. Results show that hand tracking interactions increase the sense of body ownership but lacks tracking reliability and haptic feedback. The virtual controller representation showcased better sense of agency and faster learning towards interactions. The sense of embodiment, engagement and motivation was found to be the highest in the high-fidelity VR *** a sub part of this work, we also showcase how distances can be measured in virtual environments especially in the stereoscopic Head Mounted Displays (HMD), VR. The virtual environment measured distances is then compared with the premeasured real-world distance for which the high correlation is observed. © 2023 IEEE.

关键词： Virtual reality

来源：评论

学校读者我要写书评

暂无评论

UATTA-EB: Uncertainty-Aware Test-Time Augmented Ensemble of BERTs for Classifying Common Mental Illnesses on Social Media Posts

arXiv

引用

arXiv 2023年

作者： Seth, Pratinav Agarwal, Mihir Dept. of Data Science and Computer Applications Manipal Institute of Technology Manipal Academy of Higher Education Manipal India

Given the current state of the world, because of existing situations around the world, millions of people suffering from mental illnesses feel isolated and unable to receive help in person. Psychological studies have shown that our state of mind can manifest itself in the linguistic features we use to communicate. People have increasingly turned to online platforms to express themselves and seek help with their conditions. Deep learning methods have been commonly used to identify and analyze mental health conditions from various sources of information, including social media. Still, they face challenges, including a lack of reliability and overconfidence in predictions resulting in the poor calibration of the models. To solve these issues, We propose UATTA-EB: Uncertainty-Aware Test-Time Augmented Ensembling of BERTs for producing reliable and well-calibrated predictions to classify six possible types of mental illnesses- None, Depression, Anxiety, Bipolar Disorder, ADHD, and PTSD by analyzing unstructured user data on Reddit. Copyright © 2023, The Authors. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Enhanced Framework for MRI Brain Tumor Recognition with Residual Learning and Intuitive Heatmap Visualization

Enhanced Framework for MRI Brain Tumor Recognition with Resi...

引用

2024 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems, ICSES 2024

作者： Velmurugan, L. Maheswari, R.K. Janani, K. Agalya, K. Poornima, M. Gayathri, K.C. Vit Bhopal University Bhopal-Indore Highway Kothrikalan Madhya Pradesh Sehore India M.A.M College of Enginnering and Technology Artificial Intelligence and Data Science Tamilnadu Trichy India Manipal institute of Technology Manipal Academy of Higher Education Department of Data science and computer applications Manipal India Sri Eshwar College of Engineering Department of Computer Science Engineering Coimbatore India Oasys Institute of Technology Department of Information Technology Tamil Nadu Trichy India Kelambakkam Chennai India

ISBN: (纸本)9798331543617

Early detection of brain tumors is crucial for effective treatment, yet it poses a significant challenge due to the rapid nature of the disease's progression. This study explores the application of ResNet, aimed at improving the accuracy of brain tumor diagnosis in magnetic resonance imaging (MRI) scans. Our model was trained on a dataset comprising 253 MRI images, which included 155 confirmed cases of tumors. The residual learning framework of ResNet effectively captures intricate features and mitigates issues such as vanishing gradients. The results indicate that the ResNet model achieved an impressive accuracy rate of 97%, surpassing traditional detection methods. We evaluated the model's performance using metrics like accuracy and F1-score, demonstrating its robustness. To improve the interpretability of our model's predictions, we utilized Grad-CAM (Gradient-weighted Class Activation Mapping), which highlighted critical areas in the MRI scans that influenced the model's decisions. Our findings confirm ResNet' s effectiveness in medical imaging and stress the importance of model interpretability in healthcare, allowing radiologists to make informed diagnostic choices. © 2024 IEEE.

关键词： Brain mapping

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：