检索结果-内蒙古大学图书馆

Performance *** Comparative Analysis of Multimodal Bilinear Pooling Fusion Approaches for Deep Learning-Based Visual Arabic-Question Answering Systems

引用

Computer Modeling in engineering & Sciences 2025年第4期143卷 373-411页

作者： Sarah M.Kamel Mai A.Fadel Lamiaa Elrefaei Shimaa I.Hassan Electrical Engineering Department Faculty of Engineering at ShoubraBenha UniversityCairo11629Egypt Computer Science Department Faculty of Computing and Information TechnologyKing Abdulaziz UniversityJeddah21589Saudi Arabia Department of Computer and Systems Engineering Faculty of Engineering and TechnologyBadr University in Cairo(BUC)Cairo11829Egypt Communication Systems Engineering Department Faculty of EngineeringBenha National UniversityObour11846QalyubiaEgypt

Visual question answering(VQA)is a multimodal task,involving a deep understanding of the image scene and the question’s meaning and capturing the relevant correlations between both modalities to infer the appropriate *** this paper,we propose a VQA system intended to answer yes/no questions about real-world images,in *** support a robust VQA system,we work in two directions:(1)Using deep neural networks to semantically represent the given image and question in a fine-grainedmanner,namely ResNet-152 and Gated Recurrent Units(GRU).(2)Studying the role of the utilizedmultimodal bilinear pooling fusion technique in the *** the model complexity and the overall model *** fusion techniques could significantly increase the model complexity,which seriously limits their applicability for VQA *** far,there is no evidence of how efficient these multimodal bilinear pooling fusion techniques are for VQA systems dedicated to yes/no ***,a comparative analysis is conducted between eight bilinear pooling fusion techniques,in terms of their ability to reduce themodel complexity and improve themodel performance in this case of VQA *** indicate that these multimodal bilinear pooling fusion techniques have improved the VQA model’s performance,until reaching the best performance of 89.25%.Further,experiments have proven that the number of answers in the developed VQA system is a critical factor that *** the effectiveness of these multimodal bilinear pooling techniques in achieving their main objective of reducing the model *** Multimodal Local Perception Bilinear Pooling(MLPB)technique has shown the best balance between the model complexity and its performance,for VQA systems designed to answer yes/no questions.

关键词： Arabic-VQA deep learning-based VQA deep multimodal information fusion multimodal representation learning VQA of yes/no questions VQA model complexity VQA model performance performance-complexity trade-off

来源：评论

学校读者我要写书评

暂无评论

Exit options sustain altruistic punishment and decrease the second-order free-riders, but it is not a panacea

引用

Science China(Information Sciences) 2025年第1期68卷 278-291页

作者： Chen SHEN Zhao SONG Lei SHI Jun TANIMOTO Zhen WANG Faculty of Engineering Sciences Kyushu University School of Statistics and Mathematics Yunnan University of Finance and Economics School of Computing Engineering and Digital Technologies Teesside University School of Artificial Intelligence OPtics and Electro Nics (iOPEN) Northwestern Polytechnical University

Why do humans punish free riders at their own expense? This behavior represents an evolutionary puzzle in human societies. This study explores the role of exit strategies in fostering altruistic punishment within evolutionary game theory. We extend the traditional prisoner's dilemma model by incorporating exiters, players who opt out for a small payoff while nullifying their opponent's payoff, and altruistic punishers who cooperate and punish non-cooperators. Our findings indicate that in well-mixed populations, exiters destabilize defection but do not promote altruistic punishment. In social networks, however,exiters enable altruistic punishment via cyclic dominance among altruistic punishers, defectors, and exiters. Notably, this cyclic dominance is sensitive to exit payoffs; adjustments can lead to cyclic dominance of non-punishing cooperators, defectors, and exiters, or to a bi-stable state between these two types of cyclic dominance. These results highlight the nuanced impact of exiters on altruistic punishment, emphasizing the need for careful incentivization of exit behavior. While exiters can support altruistic punishment in networked populations, their effectiveness is not a panacea and is highly sensitive to exit payoffs, indicating limits to the voluntary participation mechanism.

关键词： evolutionary game theory altruistic punishment coexistence cyclic dominance bi-stable

来源：评论

学校读者我要写书评

暂无评论

Prediction of Pressure Coefficient for SACCON Wing using Artificial Neural Network

引用

Journal of Aeronautics, Astronautics and Aviation 2025年第3期57卷 403-414页

作者： Rahman, Fahmi Izzuddin Abdul Mat, Shabudin Radzi, Nor Haizan Mohd Nasir, Nazri Kasim, Khushairi Amri Institute of Vehicle Systems and Engineering Universiti Teknologi Malaysia Johor Bahru Malaysia Faculty of Computing Universiti Teknologi Malaysia Johor Bahru Malaysia

This study explores the application of artificial neural networks (ANN) for predicting the coefficient of pressure (CP), an essential parameter in aerodynamic studies, on the SACCON wing. Traditional methods such as computational fluid dynamics (CFD) and wind tunnel tests have several drawbacks, including high computational cost, time consumption, and inconsistent data. To address these issues, this study proposes a multi-layer perceptron (MLP) neural network as an alternative. The training data for the ANN is derived from previously conducted wind tunnel experiments on the SACCON wing. The network's structure includes three hidden layers and employs the Levenberg-Marquardt backpropagation algorithm for weight updates. Data pre-processing involves normalization, and the dataset is split into training, testing, and validation sets. The results indicate that the ANN can capture general trends in CP data. However, it struggles with detailed and irregular predictions specifically in predicting tip vortex and suction peak, leading to underfitting, partly due to a lack of sufficient training data at location where vortex interaction occurred. The optimal network configuration includes a learning rate and regularization parameter of 1 x 10-5 and hidden layers containing 10, 10, and 3 neurons, respectively. Despite achieving reasonable performance metrics (RMSE = 0.1555, R-value = 0.9855, R2 = 0.9711), the network's predictions on new data are less accurate due to underfitting. Future improvements suggest increasing the training dataset size or enhancing feature engineering to improve the ANN's predictive capabilities. Additionally, using ANNs during the testing phase can reduce time, manpower, costs, and the number of data points required to define aerodynamic performance. © 2025 The Aeronautical and Astronautical Society of the Republic of China. All rights reserved.

关键词： Vortex flow

来源：评论

学校读者我要写书评

暂无评论

Optimizing Image Retrieval in Cloud Servers with TN-AGW: A Secure and Efficient Approach

引用

Journal of The Institution of Engineers (India): Series B 2025年第2期106卷 459-473页

作者： Ponnuviji, N.P. Nirmala, G. Kokila, M. L. Sworna Priyadharshini, S. Indra Department of Computer Science and Engineering RMK College of Engineering and Technology Puduvoyal Thiruvallur India Department of Computer Science and Engineering RMD Engineering College Chennai India Department of Computing Technologies School of Computing Faculty of Engineering and Technology SRM Institute of Science and Technology Tamil Nādu Kattankulathur Chennai India School of Computer Science and Engineering Vellore Institute of Technology Chennai India

The increasing use of cloud-based image storage and retrieval systems has made ensuring security and efficiency crucial. The security enhancement of image retrieval and image archival in cloud computing has received considerable attention in transmitting data and ensuring data confidentiality among cloud servers and users. Various traditional image retrieval techniques regarding security have developed in recent years but they do not apply to large-scale environments. This paper introduces a new approach called Triple network-based adaptive grey wolf (TN-AGW) to address these challenges. The TN-AGW framework combines the adaptability of the Grey Wolf Optimization (GWO) algorithm with the resilience of Triple Network (TN) to enhance image retrieval in cloud servers while maintaining robust security measures. By using adaptive mechanisms, TN-AGW dynamically adjusts its parameters to improve the efficiency of image retrieval processes, reducing latency and utilization of resources. However, the image retrieval process is efficiently performed by a triple network and the parameters employed in the network are optimized by Adaptive Grey Wolf (AGW) optimization. Imputation of missing values, Min–Max normalization, and Z-score standardization processes are used to preprocess the images. The image extraction process is undertaken by a modified convolutional neural network (MCNN) approach. Moreover, input images are taken from datasets such as the Landsat 8 dataset and the Moderate Resolution Imaging Spectroradiometer (MODIS) dataset is employed for image retrieval. Further, the performance such as accuracy, precision, recall, specificity, F1-score, and false alarm rate (FAR) is evaluated, the value of accuracy reaches 98.1%, the precision of 97.2%, recall of 96.1%, and specificity of 917.2% respectively. Also, the convergence speed is enhanced in this TN-AGW approach. Therefore, the proposed TN-AGW approach achieves greater efficiency in image retrieving than other existing

关键词： Extraction

来源：评论

学校读者我要写书评

暂无评论

The impact of backbone selection in YOLOv8 models on brain tumor localization

引用

Iran Journal of Computer Science 2025年 1-23页

作者： Ranjbarzadeh, Ramin Crane, Martin Bendechache, Malika School of Computing Faculty of Engineering and Computing Dublin City University Dublin Ireland ADAPT Research Centre School of Computer Science University of Galway Galway Ireland

This study investigates the utilization of the You Only Look Once (YOLOv8) deep learning framework for accurately identifying the location of brain tumors in medical imaging. We investigate the effects of model size and pretraining on the accuracy and computational efficiency of tumor detection by utilizing different setups of the YOLOv8 model. These setups include various configurations, ranging from very small to large, and can be pretrained on the COCO dataset or not. The experimental results, carried out on Google Colaboratory using NVIDIA Tesla T4 GPUs, show that pretrained models often achieve better performance by utilizing the extensive feature representations learned from the COCO dataset, resulting in increased precision in tumor location. For instance, the YOLOv8-XS model pretrained on COCO achieves an IoU of 0.278 and a Dice coefficient of 0.435, whereas its non-pretrained counterpart attains only 0.241 IoU and 0.388 Dice, indicating a 15% improvement in tumor localization accuracy. Similarly, pretrained YOLOv8-L achieves 0.269 IoU, outperforming standard object detection models such as Mask R-CNN (IoU: 0.212) and Faster R-CNN (IoU: 0.228). These results highlight the impact of pretraining on model performance, particularly for lightweight architectures, while also revealing diminishing returns for larger models. The research uncovers a subtle connection between the complexity of the model, pretraining, and the time required for training. It emphasizes the possible advantages and constraints of pretraining for various sizes of models. © The Author(s) 2025.

关键词： Brain tumor localization COCO dataset Deep learning Object detection Transfer learning

来源：评论

学校读者我要写书评

暂无评论

Qualitative Comparison of Vision and Vision-Language Models in Object Detection: Capabilities and Challenges 29

Qualitative Comparison of Vision and Vision-Language Models ...

引用

29th International Conference on Information Technology, IT 2025

作者： Rakic, Branka Dejanovic, Igor University of Novi Sad Faculty of Technical Sciences Department of Computing and Control Engineering Serbia

ISBN: (纸本)9798331517649

Advancements in deep learning and computer vision have greatly enhanced object detection, playing a vital role in applications such as autonomous driving and surveillance. This study explores the trade-offs between two distinct YOLO model architectures: the vision-only models, referred to as pre-YOLO-World, and a vision-language model known as YOLO-World. While the pre-YOLO-World models are renowned for their speed and accuracy, the YOLO-World model integrates large language models to enhance contextual comprehension, enabling open-vocabulary object detection. This paper provides a qualitative analysis of their strengths and limitations, focusing on speed, accuracy, adaptability, computational requirements, and application scenarios. By contrasting these architectures, we aim to offer insights into their potential roles in evolving object detection paradigms and to guide practitioners in selecting models suited to specific use cases. © 2025 IEEE.

关键词： Machine vision

来源：评论

学校读者我要写书评

暂无评论

Revolutionizing E-Commerce Customer Service: Leveraging GPT-Powered AI Chatbots to Bridge the User Expectation Gap

Revolutionizing E-Commerce Customer Service: Leveraging GPT-...

引用

International IoT, Electronics and Mechatronics Conference, IEMTRONICS 2024

作者： Mohamed, Azza Ismail, Ibrahim Faculty of Engineering and Computing IT Department Liwa College Al Ain United Arab Emirates

ISBN: (纸本)9789819747832

The use of advanced AI technologies, such as ChatGPT, has transformed customer service methods in e-commerce, particularly in response to the exceptional issues provided by the COVID-19 pandemic. These AI-powered chatbots have emerged as critical tools, enabling live conversations to quickly resolve consumer concerns, saving time, and lowering operating expenses. Despite their utility, there is still a significant gap between user expectations and the capabilities of contemporary chatbot systems. Despite significant progress toward more human-like interactions, some users are dissatisfied, leading to the view of AI as obtrusive rather than beneficial. This paper will look at how chatbots can help with customer care in the e-commerce industry. The study proposes a revolutionary chatbot solution that is focused on meeting customer expectations and improving the online buying experience. The study uses GPT-powered AI virtual chatbots to highlight the potential benefits of greatly improving the quality of customer service responses, bridging the gap between user expectations and chatbot systems’ current capabilities. The study's findings show a significant increase in customer happiness and engagement following the introduction of GPT-powered AI virtual chatbots. A review of user feedback and data revealed that the chatbot system addressed customer inquiries more accurately and efficiently, resulting in a shorter response time and higher resolution rates. Furthermore, the addition of more conversational and contextually relevant interactions improved the user experience, encouraging a good perception of AI-powered customer care among e-commerce customers. These findings highlight the great potential for GPT-powered AI chatbots to improve customer service standards and fulfill changing user expectations in the e-commerce sector. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Marketplaces

来源：评论

学校读者我要写书评

暂无评论

Performance investigation of MEA-1DMA2P solution for CO2 capture: A rate-based modelling, optimization, and cost analysis

引用

Journal of Industrial and engineering Chemistry 2025年 147卷 179-200页

作者： Afkhamipour, Morteza Shamsi, Mohammad Seifi, Ebad Hekmatiyan, Abbas Ansarpour, Meisam N.Borhani, Tohid National Iranian Gas Company South Pars Gas Complex Asaluyeh Iran Process Engineering Department Faculty of Chemical Engineering Tarbiat Modares University Tehran Iran Department of Chemical Engineering Faculty of Engineering Arak University Arak Iran School of Engineering Centre for Engineering Innovation and Research School of Engineering Computing and Mathematical Sciences University of Wolverhampton United Kingdom

This study presents a framework for evaluating the performance of MEA-1DMA2P solution in CO2 absorption for a structured packed bed by integrating a rate-based model with statistical optimization. The developed model is validated with experimental data. Sensitivity analysis identified mass transfer correlations as critical factors influencing CO2 concentration profiles, with kinetic models having a less significant impact. Operating parameters showed that CO2 removal efficiency increases with higher amine flow rate, column pressure, and amine temperature but decreases with higher CO2 gas concentration. Efficiency notably declines when the amine temperature exceeds 333 K. Statistical optimization was used to analyze the combined influence of operating parameters on CO2 removal performance. The response values for statistical optimization were obtained using the rate-based model, identifying the ranking of parameter impacts: amine flow rate had the most significant influence, followed by CO2 partial pressure, blended amine temperature, and CO2 loading of blended amine. Cost analysis was also done by scaling up the absorption system using an MEA-1DMA2P solution. The cost of CO2 capture based on the absorber column was estimated as 40.35 $/tonne of CO2, which is acceptable compared to the literature and other solvents and processes for CO2 capture. © 2024 The Korean Society of Industrial and engineering Chemistry

关键词： Sensitivity analysis

来源：评论

学校读者我要写书评

暂无评论

Development and Evaluation of a University Chatbot Using Deep Learning: A RAG-Based Approach 8th

Development and Evaluation of a University Chatbot Using Dee...

引用

8th International Workshop on Chatbots and Human-Centered AI, CONVERSATIONS 2024

作者： Olawore, Kabir McTear, Michael Bi, Yaxin Faculty of Computing Engineering and the Built Environment Ulster University Belfast Campus Belfast United Kingdom

ISBN: (纸本)9783031880445

In university systems, traditional methods of information retrieval are often found to be inefficient, leading to frustration among students and staff. This paper presents the development and evaluation of a university-specific chatbot that employs the Retrieval-Augmented Generation (RAG) approach to improve the accuracy and relevance of its responses. Unlike conventional chatbots that depend on intent classification and pre-designed system responses and conversation flows, the proposed chatbot integrates Large Language Models (LLMs) with local university data, enhancing its ability to handle complex queries with context-aware responses and dynamically generated conversation flows. The system architecture includes components such as LangChain for orchestration, a vector store for embedding external knowledge, and a user interface developed using Streamlit. Evaluation results demonstrate that the RAG-based chatbot substantially outperforms traditional LLMs, including GPT-3.5, GPT-4 mini, and GPT-4, in terms of answer accuracy and reliability. In this paper we also reflect on the lessons learned during the chatbot’s development and deployment in a real-world university setting. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Information retrieval

来源：评论

学校读者我要写书评

暂无评论

Analysis of User Experience and Task Performance in a Multi-User Cross-Reality Virtual Object Manipulation Task 17

Analysis of User Experience and Task Performance in a Multi-...

引用

17th International Workshop on IMmersive Mixed and Virtual Environment Systems, MMVE 2025, co-located with ACM Multimedia Systems Conference, MMSys 2025

作者： Brzica, Lea Matanović, Filip Vlahović, Sara Pavlin-Bernardić, Nina Skorin-Kapov, Lea University of Zagreb Faculty of Electrical Engineering and Computing Zagreb Croatia University of Zagreb Faculty of Humanities and Social Sciences Zagreb Croatia

ISBN: (纸本)9798400714689

This paper presents a user study (N=40) exploring the impact of different virtual object manipulation techniques on user experience and task performance in a networked cross-reality application. Participants were organized into pairs, with one situated in an Augmented Reality (AR) environment in a real-world lab and the other in its corresponding Virtual Reality (VR) digital twin. The task required participants to take turns moving chess pieces to predetermined positions on a chessboard. The study was organized into six scenarios, each combining one of three manipulation techniques with one of two chess set sizes. Task performance measures included completion time and error rates, while subjective measures encompassed overall Quality of Experience (QoE), usability, co-presence, quality of interaction, and consistency with expectations. Results demonstrated a clear preference for the Pick, Drag, and Snap (PDaS) manipulation technique, which showed significantly faster completion times and higher user satisfaction. This research contributes to the growing body of knowledge on designing effective interaction techniques for cross-reality systems, revealing differences in results and preferences between AR and VR groups, particularly in task completion time. © 2025 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： Virtual environments

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：