检索结果-内蒙古大学图书馆

5th International Symposium on computer engineering and Intelligent Communications, ISCEIC 2024

作者： Gu, Tao Zhang, Chongyang School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing China

ISBN: (纸本)9798331518677

Recent advancements in Vision-Language Pre-training (VLP) techniques have greatly improved performance in Scene Text Detection tasks by leveraging the rich visual and textual content in scene text images. We propose an innovative integration of contrastive learning with Masked Language Modeling (MLM) and Masked Image Modeling (MIM), inspired by the Masked Autoencoder (MAE) approach. This synthesis enhances self-supervised learning by combining discriminative and generative learning. Notably, we introduce masked image modeling for text detection, enabling effective representation learning in text image regions. To distinguish our approach from traditional BERT's MLM, we develop a custom tokenizer tailored for scene text detection. Our pre-trained model aligns visual and textual information, resulting in improved performance of existing scene text detectors. Extensive experiments on datasets such as ICDAR2015 show that our method significantly outperforms previous pre-training approaches. © 2024 IEEE.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

The Development of Polycystic Ovary Syndrome Risk Evaluation System using Advanced Machine Learning Technique 7

The Development of Polycystic Ovary Syndrome Risk Evaluation...

引用

7th International Conference on Inventive Computation Technologies, ICICT 2024

作者： Velvizhi, R. Arya Kiruthiknivas, R. Sathyabama Institute of Science And Technology Department of Computer Science And Engineering Chennai India

ISBN: (数字)9798350359299

ISBN: (纸本)9798350359299

A complex endocrine disorder that affects fertile women, polycystic ovary syndrome, also known as PCOS, is characterized by an extensive spectrum of symptoms. This work extends previous efforts in PCOS detection and tracking with the goal of utilizing state-of-the-art machine learning techniques to overcome existing constraints. The goal of the project is to improve medical outcomes by utilizing predictive analytics to analyze big patient datasets that include genetic and clinical data. In this work, robust prediction models are developed by applying machine learning techniques, including Random Forest, decision trees, and support vector machines. These models aim to accurately identify individuals who are susceptible to PCOS and forecast the course of the disease in those who have already been diagnosed. Crucially, an extensive performance assessment is provided, contrasting the suggested models with current ones to emphasize their advantages and shortcomings. In-depth dataset investigation, rigorous model training, and comprehensive assessment metrics are conducted as part of the experiments. The outcomes highlight the usefulness of the models and indicate how they could assist medical practitioners in making defensible decisions. Predictive analytics integrated into electronic health records (EHRs) allows for ongoing PCOS patient monitoring, enabling timely interventions, individualized treatment plans, and better patient results. This research advances the area by improving the accuracy of PCOS diagnosis and therapy and by addressing the shortcomings of the previous literature. The suggested machine learning methods improve patient satisfaction, health outcomes, and system productivity, in addition to enhancing identification accuracy. This revised strategy guarantees a more impactful and targeted portrayal of the research activity, in line with the input that was given. © 2024 IEEE.

关键词： Decision trees

来源：评论

学校读者我要写书评

暂无评论

BiMoeFormer: A Novel Transformer-Based Framework for 3D Human Motion Prediction with Bi-Level Routing Attention and Mixture of Experts 5

BiMoeFormer: A Novel Transformer-Based Framework for 3D Huma...

引用

5th International Seminar on Artificial Intelligence, Networking and Information Technology, AINIT 2024

作者： Zhang, Wei Nanjing University of Science and Technology School of Computer Science and Engineering Nanjing China

ISBN: (纸本)9798350385557

We introduce a novel BiMoeFormer based on the Transformer architecture for 3D human motion prediction. Previous approaches primarily focus on the relationships between body joints in human poses, while neglecting their independencies. Our method addresses this by incorporating a Bi-Level Routing Attention (BRA) mechanism, which emphasizes balancing the correlation and independence among these joints. More specifically, BRA selects the top k most relevant joints for each joint, enhancing the overall modeling capability. Moreover, to accurately capture the diversity and complexity of human motion, we employ a Mixture of Experts (MoE) architecture to enhance the expressiveness of the Transformer. This allows us to control the model complexity by sharing the MoE across blocks. Comprehensive evaluations on the Human3.6M, AMASS, and 3DPW datasets demonstrate that our method consistently outperforms state-of-the-art baselines. © 2024 IEEE.

关键词： Motion estimation

来源：评论

学校读者我要写书评

暂无评论

DeepRetina: A Deep Learning Approach for Diabetic Retinopathy Detection and Stage Classification

DeepRetina: A Deep Learning Approach for Diabetic Retinopath...

引用

2024 International Conference on Advances in Data engineering and Intelligent Computing Systems, ADICS 2024

作者： Nithishvikraman, H. Sudharsan, G. Anand Joseph Daniel, D.I. Hindustan Institute Of Technology And Science Department Of Computer Science And Engineering Chennai India

ISBN: (纸本)9798350364828

Using color fundus imaging to identify diabetic retinopathy (DR) is a tough process that requires skilled doctors to comprehend the existence and significance of certain small characteristics. This effort is further complicated by a complex categorization system. This research aims to use convolution neural network (CNN) to consistently diagnose diabetic retinopathy and grade patients into five groups or stages. An automatically generated diagnostic may be provided by a data-enhanced CNN architecture network that can identify the complex components involved in the class task, as well as exudates, hemorrhages, and micro-aneurysms in the retina, without requiring human input. This work trained CNN using data that was made accessible to the public. When compared to other algorithms on the same dataset, it demonstrated an astounding performance. It was used to the global dataset and obtained a great accuracy of 97% on the validation image. © 2024 IEEE.

关键词： Complex networks

来源：评论

学校读者我要写书评

暂无评论

Hybrid Artificial Neural Network with Advanced Noctiluca Optimization for Activity Monitoring and Prognostic Study of Signals Using EEG Signals

Hybrid Artificial Neural Network with Advanced Noctiluca Opt...

引用

2023 Winter Summit on Smart Computing and Networks, WiSSCoN 2023

作者： Kalaivani, K. Kala, R. Arun, R. Brindha, B. Thirumangaialwar, E. Babu Sri Eshwar College of Engineering Department of Computer Science and Engineering Coimbatore India Karpagam College of Engineering Department of Computer Science and Engineering Coimbatore India Hindusthan Institute of Technology Department of Computer Science and Engineering Coimbatore India Rvs College of Engineering and Technology Department of Computer Science and Engineering Coimbatore India

ISBN: (纸本)9798350334050

Before earlier, the primary reason why academics were interested in human observations from books was so they could quickly identify human cognition. Conversely, these signals from linguistic processing can also help with machine learning (ML) algorithms. The potential of using EEG cognitive function for this function is mainly untested at this time. This paper presents the first comprehensive assessment of the capability of EEG brainwave activity data for monitoring and predicting emotional analysis by analysis of recorded Electroencephalography (EEG) signals using a novel Optimized Artificial Neural network (OANN) approach. With ANN and ANO working together, we can pick the best possible values for weights. To further improve activity prediction, the proposed method imports EEG data from the Bern-Barcelona EEG collection. With an accuracy rate of 97% for OANN for emotion recognition challenges, experimental evidence confirms the efficacy of these networks in terms of effectiveness metrics. © 2023 IEEE.

关键词： Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

High Degree Dependency Study of Ensemble Models in Credit Card Fraud Analysis 2

High Degree Dependency Study of Ensemble Models in Credit Ca...

引用

2nd International Conference on Computational Intelligence, Communication Technology and Networking, CICTN 2025

作者： Rani, Shalini Tiwari, Mayank Tyagi, Shiva Department of Computer Science and Engineering Ajay Kumar Garg Engineering College Ghaziabad India

ISBN: (纸本)9798331530389

Credit card fraud detection is an increasingly critical issue due to the growth of digital transactions and the sophistication of fraudulent activities. This study proposes a hybrid framework combining Graph Neural Networks (GNNs), ensemble methods (e.g., XGBoost, Random Forest), and polynomial feature engineering to address key challenges, including class imbalance, non-linear data relationships, and real-time applicability. The system achieved a detection accuracy of 96.7%, precision of 96.9%, recall of 96.4%, and a false-negative rate of 3.6% on benchmark datasets. These improvements are attributed to the integration of advanced sampling techniques, polynomial feature transformations, and the hybrid architecture. The findings demonstrate the effectiveness of the proposed approach in enhancing fraud detection systems and provide a scalable solution for real-world applications. © 2025 IEEE.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

A Hybrid Deep Learning for Animal Species Classification using CCTV Surveillance Video: A Review 4

A Hybrid Deep Learning for Animal Species Classification usi...

引用

4th IEEE International Conference on ICT in Business Industry and Government, ICTBIG 2024

作者： More, Rahul P. Khan, Rais Abdul Hamid Sandip University School of Computer Science & Engineering Nashik India

ISBN: (纸本)9798331518981

Biodiversity is crucial for maintaining ecosystem stability, yet global biodiversity is currently in sharp decline, necessitating urgent protective measures. Wildlife monitoring and conservation, which determine biodiversity patterns, are fundamental in ecology, biogeography, and conservation biology. Advances in digital technologies, such as Closed Circuit Television (CCTV) systems, have enhanced the ability to monitor wildlife activities. The growth in wildlife data has facilitated the detection of targeted wild animals and their behaviors, but challenges remain due to the continuous data stream and complex environments. Recent developments in Deep Learning (DL) have demonstrated potential in overcoming these challenges by efficiently processing large-scale video data to detect and classify animal activities. Despite these advancements, further research is needed to refine these models for real-time applications, especially in mitigating wildlife-human conflicts and ensuring the safety of both wildlife and humans. This paper explores the integration of DL techniques in wildlife monitoring, highlighting their potential to improve biodiversity conservation efforts. In this review article we overview the deep learning approaches, focusing on detection of wild creatures to lessen the wild animal- human conflicts. We present a comprehensive overview of issues and challenges in wild life detection in order for a better understanding. © 2024 IEEE.

关键词： Invertebrates

来源：评论

学校读者我要写书评

暂无评论

Sentiment Analysis of Financial News: Implications for Corporate Strategy in Risk Management, Mergers and Marketing

Sentiment Analysis of Financial News: Implications for Corpo...

引用

2025 International Conference on Multi-Agent Systems for Collaborative Intelligence, ICMSCI 2025

作者： Sendhil, R. Lijo, Joshua Abinaya, E. School of Computer Science and Engineering Vellore Institute of Technology Chennai India

ISBN: (纸本)9798331509828

This research study examines the ways to use sentiment analysis on financial news for corporate strategy making. We examine the impact of sentiment in financial news on corporate decisions (beyond technology) regarding risk management, mergers and marketing strategy using recent developments from research into sentiment analysis as input and contrasts it with our findings. This study presents unsupervised and supervised sentiment analysis methods, statistical methods and AI tools for sentiment analysis, and their contributions to identifying business needs. In addition, the paper underscores existing difficulties and received a number of limitations on current techniques in sentiment analysis that have to be tended for better supporting corporate strategy. © 2025 IEEE.

关键词： Decentralized finance

来源：评论

学校读者我要写书评

暂无评论

GSVAELP: integrating graphSAGE and variational autoencoder for link prediction

引用

Multimedia Tools and Applications 2024年 1-27页

作者： Ziya, Fatima Kumar, Sanjay Department of Computer Science and Engineering Delhi Technological University Main Bawana Road Delhi New Delhi110042 India

Link prediction (LP) plays a crucial role in network science, which forecasts potential connections or relationships between nodes or entities within the network. Link prediction has found many applications, such as suggesting new connections to users in social networks, personalized recommendations in e-commerce, predicting new routes in transportation networks, and many others. This paper introduces a deep learning-based link prediction model in social networks by leveraging graphSAGE (graph sample and aggregation) and Variational Autoencoders (VAE). The proposed work starts by utilizing a graphSAGE to generate node embeddings of the input network by sampling and aggregating information from neighborhood nodes. The generated embeddings are sufficiently expressive to capture the local and global network structure. Further, we adopt VAE to learn a latent space representation of the graphSAGE embeddings. The VAE helps to refine the node embeddings and learn a meaningful latent space representation of the input data, which can be useful for the downstream link prediction task. The encoder’s output (latent space) can capture important features that aid link prediction. Finally, we train a logistic regression classifier using the latent representations from the VAE as features to predict the upcoming links in the network. The necessary hyperparameter studies are performed to obtain the optimal values of the various model parameters. The experiments and simulations conducted on eight different real-world network datasets illustrate the effectiveness of the proposed link prediction model. Additionally, we evaluate the average performance of each comparative link prediction method across all datasets to assess the efficacy of the proposed approach. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

Sentiment Classification for Depression Detection: Integrating Capsule Networks with CNNs on Review Data

Sentiment Classification for Depression Detection: Integrati...

引用

2025 International Conference on Emerging Technologies for Intelligent Systems, ETIS 2025

作者： Sagili, Sreenivas Reddy Shibi, B. Rvs, Praveen Anjana, P. Rochester Institute of Technology Department of Computer Science United States Sona College of Technology Computer Science and Engineering Tamil Nadu Salem India LTI Mindtree Digital Engineering India Awh Engineering College Computer Science and Engineering Kerala India

ISBN: (纸本)9798331507541

This research is focused on the analysis of how Capsule Networks should be combined with Convolutional Neural Networks (CNNs) for improving the sentiment classification with the target of depression detection based on the data offered by the reviews. Other methodologies of analysing the sentiment of a phrase may encounter disability in the process of analysing texts in natural language following depression since they do not analyse the text deeply enough to consider the context. The proposed method intends to enhance accuracy and resistance when compared to state-of-the-art techniques by combining the attributes of Capsule Networks, that effectively maintain the spatial hierarchy and the connections within the data, with CNNs' feature extraction ability. The proposed hybrid model helps overcome disadvantages of the plain sentiment analysis because it identifies intricate patterns and dependencies in the review data and enhances the depressive sentiments' detection. This paper assesses the propriety of the integrated Capsule-CNN through an evaluation on review datasets in relation to its ability in the successful differentiate between depressive and non-depressive sentiments. The findings reveal a considerable increase in method accuracy and present a leap of progress in the field of automatic depression detection, which in turn helps to develop more accurate and efficient tools for monitoring patients' mental status. © 2025 IEEE.

关键词： Emotion Recognition

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：