检索结果-内蒙古大学图书馆

1st Workshop on Writing Aids at the Crossroads of AI, Cognitive science and NLP, WRAICOGS 2025

作者： Shi, Ken Penn, Gerald Dept. of Computer Science University of Toronto Canada

ISBN: (纸本)9798891762213

In this paper, we introduce the concept of Semantic Masking, where semantically coherent surrounding text (the haystack) interferes with the retrieval and comprehension of specific information (the needle) embedded within it. We propose the Needle-in-a-Haystack-QA Test, an evaluation pipeline that assesses LLMs’ long-text capabilities through question answering, explicitly accounting for the Semantic Masking effect. We conduct experiments to demonstrate that Semantic Masking significantly impacts LLM performance more than text length does. By accounting for Semantic Masking, we provide a more accurate assessment of LLMs’ true proficiency in utilizing extended contexts, paving the way for future research to develop models that are not only capable of handling longer inputs but are also adept.at navigating complex semantic landscapes. © WRAICOGS 2025 - Writing Aids at the Crossroads of AI, Cognitive science and NLP, Proceedings of the 1st Workshop.

关键词： Needles

来源：评论

学校读者我要写书评

暂无评论

Non-Intrusive Alert For Drivers With Fatigue 4

Non-Intrusive Alert For Drivers With Fatigue

引用

4th IEEE Asian Conference on Innovation in Technology, ASIANCON 2024

作者： Amrutha, K. Shreekumar, T. Resmi, K.R. Vineetha, K.R. Dept. of Computer Science Bangalore India Mite Dept of Computer Science and Engineering Mangalore India

ISBN: (数字)9798350354218

ISBN: (纸本)9798350354218

Fatigue during driving significantly impairs a driver's reaction time, awareness, and decision-making abilities, leading to an increased risk of accidents. Recognising and mitigating driver fatigue is crucial for enhancing road safety and preventing potential collisions. This research introduces a non-intrusive method to reduce road accidents caused by driver fatigue. Recognising that accidents often occur due to lapses in driver alertness, the system utilises an affordable, high-quality webcam and dashcam. The approach involves extracting essential information through a multi-stage monitoring process to evaluate the driver's state thoroughly. The study includes a detailed analysis of webcam specifications, such as resolution, frame rate, and additional features. It also compares the traditional cascading technique with the advanced MediaPipe method, incorporating facial recognition, eye movement tracking, and head position analysis to detect driver fatigue accurately. © 2024 IEEE.

关键词： Eye movements

来源：评论

学校读者我要写书评

暂无评论

A comprehensive step of XAI for Bengali and English text detection and recognition from natural scene images 2

A comprehensive step of XAI for Bengali and English text det...

引用

2nd International Conference on Networking, Embedded and Wireless Systems, ICNEWS 2024

作者： Dutta, Mithun Das, Utpol Kanti Das, Jugal Krishna Rangamati Science and Technology University Dept. of Computer Science and Engineering Rangamati Bangladesh Jahangirnagar University Dept. of Computer Science and Engineering Dhaka Bangladesh

ISBN: (纸本)9798350367461

A critical first step in many applications, including augmented reality, document analysis, and scene comprehension, is text detection from images. Even though text identification for the English language has advanced significantly, there are particular difficulties when applying these methods to languages such as Bengali because of variations in script, morphology, and typographic characteristics. This article offers a wonderful method for leveraging the Faster R-CNN framework to identify and categorize text from images that contain Bengali and English sentences. The study utilizes a dataset named "Bengali and English text image dataset (BETID)"that consists of images with text in both Bengali and English scripts, taken from a variety of sources such as books, documents, natural scenes, internet images, and landscapes. Pre-processing methods are used to improve the resilience and generalization of the model, including image scaling, normalization, and augmentation. The Faster R-CNN architecture is tuned and trained to identify text regions in images and categorize Bengali or English text. The performance of the text Classification model is evaluated using evaluation measures such as accuracy on the BETID dataset. The proposed Faster R-CNN model achieved an average of 97.90% accuracy, which is better than other deep learning models. Empirical findings showcase the capacity of transfer learning to alleviate data scarcity concerns and enhance text detection precision for low-resource languages such as Bengali.) © 2024 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Comparative analysis of Machine Learning Techniques for the Detection of Heart Disease

Comparative analysis of Machine Learning Techniques for the ...

引用

2024 International Conference on science, Engineering and Business for Driving Sustainable Development Goals, SEB4SDG 2024

作者： Adeniyi, Jide Kehinde Adeniyi, Tunde Taiwo Adeniyi, Emmanuel Abidemi Abiodun, Moses Kazeem Awotunde, Joseph Bamidele Landmark University SDG 9: Industry Innovation and Infrastructure Dept. of Computer Science Nigeria University of Ilorin Dept. of Computer Science Nigeria Bowen University Dept. of Computer Science Iwo Nigeria Landmark University SDG 10: Reduced Inequality Dept. of Computer Science Nigeria

ISBN: (纸本)9798350358155

The World Health Organization stated that an alarming 32 percent of worldwide fatalities was from heart related challenges. This percentage is even more escalated in poor and moderate countries. Most heart related issues can be ameliorated or managed if they are detected on time. Hence, with the advent of machine learning techniques, prediction of heart related issues can be made. Several machine learning methods have been proposed for heart related issues prediction, and each with varying performance level. A few comparisons have been made amongst these machine learning techniques with little details. In this study, a detailed comparison of four machine learning techniques was considered. The machine learning methods considered were selected based on their performance from literatures. The examined methods include the decision tree, random forest, support vector machines, and logistic regression. Prior to training and testing, the dataset was preprocessed by encoding the categorical data and scaling the dataset features. The outcome of the testing showed that the accuracy was obtained for logistic regression and support vector machine;with both having an accuracy of 0.882. © 2024 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

State of the art: Security Testing of Machine Learning Development Systems 14

State of the art: Security Testing of Machine Learning Devel...

引用

2024 IEEE 14th Annual Computing and Communication Workshop and Conference, CCWC 2024

作者： Das, Saikat Krishnamurthy, Bhargavi Das, Raktim Ranjan Shiva, Sajjan G Utah Valley University Dept. of Computer Science Orem United States Siddaganga Institute of Technology Dept. of Computer Science Karnataka Tumakuru India Stamford University Bangladesh Dept. of Computer Science Dhaka Bangladesh University of Memphis Dept. of Computer Science Memphis United States

ISBN: (纸本)9798350360134

In recent days, machine learning (ML) systems have become integral to nearly all mainstream applications. Understanding the underlying logic that contributes to the desired behavior in ML systems can be challenging. Humans play a crucial role in providing the necessary data samples for training ML models to ensure accurate predictions. The complexity of ML testing arises from various factors such as data dependency, dynamic model behavior, the absence of a test oracle, a vast input space, and the lack of a specific testing life cycle tailored for ML. Testing ML systems is not a straightforward process, as it not only involves verifying the code and their corresponding data but also is further complicated by the dynamic nature of these systems. A typical ML model undergoes seven stages, including eliciting business needs/requirements, gathering data, selecting a model, training the model, testing the model, deploying the model, and monitoring the model. Throughout the ML Development Life Cycle (MLDLC), each stage introduces various security risks. This paper explores the current state of the literature on security attacks and defense approaches concerning data, model, and prediction output of ML models. The paper also addresses potential security attacks at each stage of the MLDLC and the corresponding security measures to mitigate those attacks. © 2024 IEEE.

关键词： Life cycle

来源：评论

学校读者我要写书评

暂无评论

A review on 3D image reconstruction on specific and generic objects

引用

Materials Today: Proceedings 2023年 80卷 2400-2405页

作者： Shalma, H. Selvaraj, P. Dept. of Computer Science and Engineering SRM Institute of Science and Technology Kattankulathur Chennai India Dept. of Information Technology SRM Institute of Science and Technology Kattankulathur Chennai India

With the emergence of various techniques involved in deep learning the researchers of computer vision tends to focus on the strategies such as object recognition and segmentation of image. This has inclined to accomplish the deep learning techniques in 3D reconstruction of both specific and generic objects. As the space for reconstruction of 3D images either in single or multi view has envisioned the researchers to concentrate on the available technologies used for reconstruction. With the available built in methods and technologies in deep learning, the performance of the proposed methods were reviewed and analyzed using several parameters. As the remaking of 2D images is still in the beginning stage, it is important to study the 3D shape representations, various network architecture, methodologies and approaches behind 3D reconstruction. In this work a review of deep learning methods for single or multiple RGB images of specific and generic object 3D reconstruction was done. Several methods and their importance were also discussed along with the challenges encountered and with further research directions. This paper critically analyzes the various 3D Shaped Representations, 3D data Network Architectures, dept. Estimation methods, Multi View Representations and the data Representation Techniques. © 2021

关键词： Object recognition

来源：评论

学校读者我要写书评

暂无评论

Cybersecurity for Industrial IoT: A Review on Vulnerabilities and Attacks 8

Cybersecurity for Industrial IoT: A Review on Vulnerabilitie...

引用

8th International Conference on Electronics, Materials Engineering and Nano-Technology, IEMENTech 2025

作者： Kumari, Reetu Kumar, Sunil Amandeep Dutta, Kamlesh Kumari, Aparna Guru Jambheshwar University of Science and Technology Dept. of Computer Science & Engineering Haryana Hisar India Guru Jambheshwar University of Science and Technology Dept. of AI & Data Science Haryana Hisar India National Institute of Technology Hamirpur Dept. of Computer Science & Engineering Himachal Pardesh India Institute of Technology Nirma University Dept. of Computer Science & Engineering Gujarat Ahmedabad India

ISBN: (纸本)9798331520762

Internet has evolved from a network of connecting people to a network of connecting things, leading to a more complex and sophisticated network of Industrial Things, known as Industrial IoT (IIoT) today. This evolution has revolutionized the traditional Industrial system into real-time smarter systems that enhanced overall efficiency and productivity. Despite its benefits, the interconnected nature of IIoT, expanded the attack surface, exposing the systems to new crucial and complex cybersecurity threats. Several real time attacks on various industries as mentioned in this paper and the increasing cybersecurity cost over the years, clearly states the need of robust security solutions. To deeply understand the cybersecurity aspect, we provided a comprehensive study of attacks and vulnerabilities in this paper. In this paper, we did a indept. study of various IIoT related severe attacks and aspects related to these attacks, like vulnerabilities, impacts, potential solutions etc. We also provided future research directions to help the research community, to work upon and develop more robust and efficient cybersecurity solutions. © 2025 IEEE.

关键词： attacks cybersecurity etc IIoT Industrial IoT IoT vulnerabilities

来源：评论

学校读者我要写书评

暂无评论

Unravelling the Enigma of Polycystic Ovary Syndrome (PCOS): Using MLAlgorithms 2

Unravelling the Enigma of Polycystic Ovary Syndrome (PCOS): ...

引用

2nd International Conference on Disruptive Technologies, ICDT 2024

作者： Juneja, Sonam Bhati, Bhoopesh Singh Atwal, Shikha Anand, Monika Maiti, Souvik Chandigarh University Dept. of Computer Science and Engineering Gharuan India Iiit Dept. of Computer Science and Engineering Sonepat India

ISBN: (纸本)9798350371055

The difficulties and ramifications of PCOS, which affects a sizable portion of women who are of reproductive age, are discussed in this review. The article covers the various clinical manifestations of PCOS, how it affects both reproductive and non-reproductive health, and how it's linked to psychological distress and metabolic disorders. It highlights how critical lifestyle modifications, early detection, and accurate diagnosis are. Additionally, the study presents machine learning techniques for PCOS detection, demonstrating the effectiveness of techniques like Random Forests, CNN, and SVM. An innovative CDSS that uses the Red Deer Algorithm shows encouraging accuracy. The necessity of continued research, diversified datasets, and cooperative efforts to enhance PCOS detection at the nexus of technology and healthcare is highlighted in the abstract's conclusion. © 2024 IEEE.

关键词： Deep Learning KNN Machine Learning PCOS Random Forest SVM

来源：评论

学校读者我要写书评

暂无评论

Architectural Model of Deep Convolution Neural Network Approach for Brain Tumor Prediction

Architectural Model of Deep Convolution Neural Network Appro...

引用

2024 International Conference on Intelligent Computing and Sustainable Innovations in Technology, IC-SIT 2024

作者： Panigrahi, Raghunath Sahu, Rekha Nayak, Suvendu Chandan Kabat, Manas Ranjan IMIT Dept. of Computer Science and Engineering Cuttack India Silicon University Dept. of Computer Science and Engineering Bhubaneswar India

ISBN: (纸本)9798350369175

Magnetic Resonance Imaging (MRI) is most commonly used technique to detect a tumor by physicians in medical science. The MRI images are three-dimensional images with the number of slices that are analyzed to identify the defective cells in the naked eye that leads to a time-consuming process. Machine Learning (ML) techniques are investigated as a solution to improve tumor identification in a better way. Early and accurate detection is crucial for effective treatment planning and improving patient outcomes. Traditionally, brain tumor diagnosis relies on manual analysis of MRI (Magnetic Resonance Imaging) scans by radiologists, which can be time-consuming and prone to human error. The advent of deep learning, particularly Convolutional Neural Networks (CNNs), has opened new avenues for automating and improving the accuracy of brain tumor prediction. In this paper, we study the Convolutional Neural Network(CNN) model which is popular in image analysis. Similar way, we develop a CNN model to process and classify the MRI images to detect the type of brain tumor. We examine the model's performance by using 5712 MRI images from a standard dataset available in the public domain. The aim of the work is to classify the tumor in four classes like: Glioma, Meningioma, Pituitary and No Tumour using the proposed CNN model. Our model follows data preprocessing, image encoding, filtering and classification. The proposed model scores 78% accuracy as its performance in the dataset which needs more investigations and comparison with other techniques in further study. © 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

A Motor Activity Analysis as a Depression Indicator: Predictive Approach Using A Hybrid CNN and LSTM Network 6

A Motor Activity Analysis as a Depression Indicator: Predict...

引用

6th International Conference on Cybernetics and Intelligent System, ICORIS 2024

作者： Sutedja, Vinsensius Odilo Andreas, Frans Nabiilah, Ghinaa Zain Moniaga, Jurike V. Dept. Computer Science School of Computer Science Bina Nusantara University Jakarta Indonesia

ISBN: (纸本)9798331539603

Depression is a prevalent mental disorder affecting behavior, cognition, and physical activity. Early detection and intervention are crucial to mitigating its severe impacts, including suicidal ideation or suicide attempts. This study proposes a novel predictive model using a hybrid Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) network to analyze motor activity data as an indicator of depression. Utilizing the "Depresjon" dataset, which comprises activity levels from both depressed and non-depressed individuals, the model extracts features through CNN and processes these with LSTM to identify depression patterns. To address class imbalance in the dataset, Synthetic Minority Over-sampling Technique (SMOTE) was employed, enhancing the model's predictive accuracy. The final model achieved a high accuracy rate of 97.89%, significantly outperforming the other model created. This research demonstrates the potential of combining CNN and LSTM for effective depression prediction, providing a valuable tool for early diagnosis and intervention. © 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：