检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Ragano, Alessandro Martinez, Helard Becerra Hines, Andrew The Insight Centre for Data Analytics The School of Computer Science University College Dublin Dublin Ireland

This study investigates the evaluation of multimedia quality models, focusing on the inherent uncertainties in subjective Mean Opinion Score (MOS) ratings due to factors like rater inconsistency and bias. Traditional statistical measures such as Pearson’s Correlation Coefficient (PCC), Spearman’s Rank Correlation Coefficient (SRCC), and Kendall’s Tau (KTAU) often fail to account for these uncertainties, leading to inaccuracies in model performance assessment. We introduce the Constrained Concordance Index (CCI), a novel metric designed to overcome the limitations of existing metrics by considering the statistical significance of MOS differences and excluding comparisons where MOS confidence intervals overlap. Through comprehensive experiments across various domains including speech and image quality assessment, we demonstrate that CCI provides a more robust and accurate evaluation of instrumental quality models, especially in scenarios of low sample sizes, rater group variability, and restriction of range. Our findings suggest that incorporating rater subjectivity and focusing on statistically significant pairs can significantly enhance the evaluation framework for multimedia quality prediction models. This work not only sheds light on the overlooked aspects of subjective rating uncertainties but also proposes a methodological advancement for more reliable and accurate quality model evaluation. © 2024, CC BY.

关键词： Image quality

来源：评论

学校读者我要写书评

暂无评论

Conformal prediction for deep classifier via label ranking 24

Conformal prediction for deep classifier via label ranking

引用

Proceedings of the 41st International Conference on Machine Learning

作者： Jianguo Huang Huajun Xi Linjun Zhang Huaxiu Yao Yue Qiu Hongxin Wei Department of Statistics and Data Science Southern University of Science and Technology and The School of Information Science and Technology ShanghaiTech University Department of Statistics and Data Science Southern University of Science and Technology Department of Statistics Rutgers University Department of Computer Science University of North Carolina at Chapel Hill College of Mathematics and Statistics Chongqing University

Conformal prediction is a statistical framework that generates prediction sets containing groundtruth labels with a desired coverage guarantee. The predicted probabilities produced by machine learning models are generally miscalibrated, leading to large prediction sets in conformal prediction. To address this issue, we propose a novel algorithm named Sorted Adaptive Prediction Sets (SAPS), which discards all the probability values except for the maximum softmax probability. The key idea behind SAPS is to minimize the dependence of the non-conformity score on the probability values while retaining the uncertainty information. In this manner, SAPS can produce compact prediction sets and communicate instance-wise uncertainty. Extensive experiments validate that SAPS not only lessens the prediction sets but also broadly enhances the conditional coverage rate of prediction sets.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Comparative Study of Deep Learning Models for Autism Diagnosis in Children Using Image Analysis

Comparative Study of Deep Learning Models for Autism Diagnos...

引用

IEEE International WIE Conference on Electrical and computer Engineering (WIECON-ECE)

作者： Mahmudul Hassan Tania Sultana Tamanna Razin Sumyta Monsoor Shehrin Hoque Rageeb Mohammad Ridwan Md. Ashraful Alam Department of Computer Science and Engineering (CSE) School of Data and Sciences (SDS) Brac University Dhaka

ISBN: (数字)9798331535476

ISBN: (纸本)9798331535483

Autism Spectrum Disorder (ASD) is a complex neurological disorder related to an individual's psychological difficulties which eventually impact their behavior o r reactions to the outside world. Identifying autism at a younger age offers several advantages, including the opportunity for the individual to lead a better life, enabling preparation for their future and that of their close family members, and contributing to increased awareness and understanding of various medical conditions. In this paper, we suggest a deep learning-based method that makes use of image datasets to identify ASD in children. The databases include facial patterns and additional visual clues that can be deduced from images. Deep learning models like VGG16, VGG19, EfficientNetB4 a nd MobileNet are used. These architectures are pretrained on large-scale image datasets and refined toe xtract discriminative features on the ASD-specific dataset. W e h ave acquired facial image datasets from a publicly available platform called Kaggle. Our primary goal is to compare the deep learning models that better fit t he dataset a nd improve t he a ccuracy of autism detection than any other work before. This paper aims to facilitate model comparisons and streamline the autism detection process using advanced deep-learning techniques available today.

关键词： Deep learning Training Autism Analytical models Visualization Psychology Streaming media Feature extraction Visual databases Testing

来源：评论

学校读者我要写书评

暂无评论

Efficient Twin Migration in Vehicular Metaverses: Multi-Agent Split Deep Reinforcement Learning with Spatio-Temporal Trajectory Generation

引用

IEEE Transactions on Mobile Computing 2025年

作者： Chen, Junlong Kang, Jiawen Xu, Minrui Wu, Fan Zhang, Hongliang Huang, Huawei Niyato, Dusit Mao, Shiwen Guangdong University of Technology School of Automation Guangzhou China Nanyang Technological University School of Computing and Data Science Singapore Singapore Central South University School of Computer Science and Engineering Changsha China Peking University School of Electronics Beijing China Sun Yat-Sen University School of Data and Computer Science Guangzhou China Auburn University Department of Electrical and Computer Engineering Auburn United States

Vehicle Twins (VTs) as digital representations of vehicles can provide users with immersive experiences in vehicular metaverse applications, e.g., Augmented Reality (AR) navigation and embodied intelligence. VT migration is an effective way that migrates the VT when the locations of physical entities keep changing to maintain seamless immersive VT services. However, an efficient VT migration is challenging due to the rapid movement of vehicles, dynamic workloads of Roadside Units (RSUs), and heterogeneous resources of the RSUs. To achieve efficient migration decisions and a minimum latency for the VT migration, we propose a multi-agent split Deep Reinforcement Learning (DRL) framework combined with spatio-temporal trajectory generation. In this framework, multiple split DRL agents utilize split architecture to efficiently determine VT migration decisions. Furthermore, we propose a spatio-temporal trajectory generation algorithm based on trajectory datasets and road network data to simulate vehicle trajectories, enhancing the generalization of the proposed scheme for managing VT migration in dynamic network environments. Finally, experimental results demonstrate that the proposed scheme not only enhances the Quality of Experience (QoE) by 29% but also reduces the computational parameter count by approximately 25% while maintaining similar performances, enhancing users' immersive experiences in vehicular metaverses. © 2002-2012 IEEE.

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

A Novel Optimization Strategies for Detecting and Preventing Distributed Denial of Service Attacks in Wireless Networks 3

A Novel Optimization Strategies for Detecting and Preventing...

引用

3rd International Conference on Smart Generation Computing, Communication and Networking, SMART GENCON 2023

作者： Kumar, Saurabh Ratna Raja Kumar, Jambi Mondal, Dipannita Hemelatha, S. Geetha, B. Diwakar, Mandar P. Mangalayatan University Department of Commerce Aligarh India Genba Sopanrao Moze College of Engineering Balewadi Department of Computer Engineering Maharashtra Pune India Dr. D. Y.Patil College of Engineering and Innovation Artificial Intelligence and Data Science Department Talegaon India Karpagam Academy of Higher Education Department of Computer Science Engineering Coimbatore641021 India Prince Shri Venkateshwara Padmavathy Engineering College Department of Science and Humanities Chennai India Vishwakarma Institute of Information Technology Department of Artificial Intelligence & Data Science Pune India

ISBN: (纸本)9798350319125

allotted Denial of carrier (DDoS) assaults have end up a prime difficulty in trendy wireless networks. these assaults aim to disrupt the regular operation of networks by means of overwhelming them with excessive traffic from more than one assets. traditional DDoS detection and prevention strategies have established to be useless in wireless networks due to the dynamic and unpredictable nature of the wireless surroundings. To deal with this problem, novel optimization techniques were proposed to enhance the potential of wireless networks to discover and prevent DDoS *** of the key techniques is the use of wise algorithms which includes machine learning and synthetic intelligence. these algorithms can analyze and study from network site visitors patterns to pick out strange behavior this is indicative of a DDoS attack. by way of continuously tracking the network and adapting to converting traffic patterns, these algorithms are capable of as it should be locate and mitigate DDoS assaults in actual-time. any other method is the deployment of robust and dynamic defense mechanisms. those mechanisms include the usage of anomaly detection, traffic filtering, and traffic shaping. Anomaly detection strategies rent statistical techniques to hit upon patterns that deviate from ordinary network conduct. traffic filtering involves figuring out and blockading malicious site visitors, even as site visitors shaping manages and controls the flow of legitimate visitors to prevent congestion and enhance community overall performance. © 2023 IEEE.

关键词： filtering legitimate ordinary robust strategies

来源：评论

学校读者我要写书评

暂无评论

Numerical study of valence band states evolution in AlxGa1-xAs QDs systems

PeerJ Materials Science

引用

PeerJ Materials science 2024年 6卷 e32-e32页

作者： Lazarev, Mikhail Faculty of Computer Science Laboratory of Methods for Big Data Analysis HSE University Moscow Russia

Quantum dots (QDs) are very attractive nanostructures from an application point of view due to their unique optical properties. Optical properties and valence band (VB) state character was numerically investigated with respect to the effects of nanostructure geometry and composition. Numerical simulation was carried out using the Luttinger–Kohn model adapted to the particular case of QDs in inverted pyramids. We present the source code of the 4-band Luttinger–Kohn model that can be used to model AlGaAs or InGaAs nanostructures. The work focuses on the optical properties of GaAs/AlGaAs [111] QDs and quantum dot molecules (QDMs). We examine the dependence of Ground State (GS) optical properties on the structural parameters and predict optimal parameters of the QD/QDM systems to achieve dynamic control of GS polarization by an applied electric field. © 2024 Lazarev.

关键词： Polarization

来源：评论

学校读者我要写书评

暂无评论

UVTM: Universal Vehicle Trajectory Modeling with ST Feature Domain Generation

引用

IEEE Transactions on Knowledge and data Engineering 2025年

作者： Lin, Yan Hu, Jilin Guo, Shengnan Yang, Bin Jensen, Christian S. Lin, Youfang Wan, Huaiyu Aalborg University Department of Computer Science Aalborg9220 Denmark East China Normal University School of Data Science and Engineering Shanghai200050 China Beijing Jiaotong University Beijing Key Laboratory of Traffic Data Mining and Embodied Intelligence School of Computer Science and Technology Beijing100044 China

Vehicle movement is frequently captured in the form of GPS trajectories, i.e., sequences of timestamped GPS locations. Such data is widely used for various tasks such as travel-time estimation, trajectory recovery, and trajectory prediction. A universal vehicle trajectory model could be applied to different tasks, removing the need to maintain multiple specialized models, thereby reducing computational and storage costs. However, creating such a model is challenging when the integrity of trajectory features is compromised, i.e., in scenarios where only partial features are available or the trajectories are sparse. To address these challenges, we propose the Universal Vehicle Trajectory Model (UVTM), which can effectively adapt to different tasks without excessive retraining. UVTM incorporates two specialized designs. First, it divides trajectory features into three distinct domains. Each domain can be masked and generated independently to accommodate tasks with only partially available features. Second, UVTM is pre-trained by reconstructing dense, feature-complete trajectories from sparse, feature-incomplete counterparts, enabling strong performance even when the integrity of trajectory features is compromised. Experiments involving four representative trajectory-related tasks on three real-world vehicle trajectory datasets provide insight into the performance of UVTM and offer evidence that it is capable of meeting its objectives. © 1989-2012 IEEE.

关键词： Travel time

来源：评论

学校读者我要写书评

暂无评论

Volume Hiding Conjunctive Searchable Symmetric Encryption over Cloud 16

Volume Hiding Conjunctive Searchable Symmetric Encryption ov...

引用

16th International Conference on Cyber-Enabled Distributed Computing and Knowledge Discover, CyberC 2024

作者： Deng, Qinghua Chen, Lanxiang Zhu, Yizhao Mu, Yi City University of Macau Faculty of Data Science 999078 China Hanshan Normal University School of Computer and Information Engineering Chaozhou521041 China Fujian Normal University Fujian Provincial Key Laboratory of Network Security and Cryptology College of Computer and Cyber Security Fuzhou China

ISBN: (纸本)9798331506896

Symmetric searchable encryption (SSE) is crucial for promoting cloud storage. It enables users to encrypt private data and outsource it to untrusted servers while retaining query capabilities. To prevent the cloud from obtaining users' private data, researchers have proposed numerous schemes to protect users' privacy. However, the issue of volume hiding has received less attention. Recent studies have demonstrated that volume leakage can be exploited to infer query *** mitigate the aforementioned leakage problem, we introduce an efficient volume-hiding SSE scheme supporting conjunctive queries, denoted as VHXT. The primary strategy for implementing volume hiding comprises two aspects: first, employing a bitmap-based inverted index to obscure the query response length of the least frequent keyword, and second, introducing laconic PSI for efficient membership testing. The laconic PSI enables our scheme to achieve approximately optimal overhead. Moreover, security analysis and performance evaluation demonstrate that VHXT effectively prevents volume leakage with minimal overhead. © 2024 IEEE.

关键词： Information leakage

来源：评论

学校读者我要写书评

暂无评论

Enhancing Citywide Flow Forecasting through Deep Learning and Spatial Temporal Learning 2

Enhancing Citywide Flow Forecasting through Deep Learning an...

引用

2nd IEEE International Conference on Integrated Circuits and Communication Systems, ICICACS 2024

作者： Parvathi, R. Sarkar, Swagata Supraja, M. Swagatha, J.P. Kanagajothi, D. Srimathi, S. SRM Arts and Science College Computer Science Kattankulathur India Sri Sairam Engineering College Department of Artificial Intelligence and Data Science Chennai India Rajarajeshwari College of Engineering Electronics and Communication Engineering Bangalore India Panimalar Engineering College Department of IT Poonamallee India Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology Department of Mathematics India SIMATS Technology Saveetha School of Engineering Department of Bio Chennai India

ISBN: (纸本)9798350317558

Safe traffic management requires citywide flow projections. Transportation, public safety, and municipal planning are substantially affected. It forecasts city intake and outflow using flow data. Traditional methods are inaccurate and less scalable in complex, dynamic metropolitan settings. This study uses deep learning to predict crow traffic. Deep learning can handle complicated nonlinear circumstances, making neural network crowd flow prediction more popular. However, deep learning has solved this problem, but it requires plenty of data. data shortages may develop when a city's service or infrastructure is installed. This research introduces a deep spatial temporal transfer learning system that predicts crowd movement in a data-sparse target city using data from a data-rich source city. Geographic and temporal patterns in crow traffic data may be captured by trained algorithms, enhancing prediction accuracy. We found that our deep learning-based solution is more accurate and robust than prior methods. Crow migration patterns are properly predicted by this research, improving urban planning and management. © 2024 IEEE.

关键词： Learning systems

来源：评论

学校读者我要写书评

暂无评论

GANs-guided Conditional Diffusion Model for Synthesizing Contrast-enhanced Computed Tomography Images 46

GANs-guided Conditional Diffusion Model for Synthesizing Con...

引用

46th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2024

作者： Yang, Yulin Liu, Jing Chen, Qingqing Li, Yinhao Han, Xian-Hua Hu, Hongjie Lin, Lanfen Chen, Yen-Wei Ritsumeikan University Gradate School of Information Science and Engineering Shiga Japan Zhejiang Lab. Research Center for Healthcare Data Science Hangzhou China Zhejiang University School of Medicine Department of Radiology Sir Run Run Shaw Hospital Hangzhou China Ritsumeikan University College of Information Science and Engineering Shiga Japan Yamaguchi University Artificial Intelligence Research Center Yamaguchi Japan Zhejiang University College of Computer Science and Technology Hangzhou China

ISBN: (纸本)9798350371499

In contrast to non-contrast computed tomography (NC-CT) scans, contrast-enhanced (CE) CT scans can highlight discrepancies between abnormal and normal areas, commonly used in clinical diagnosis of focal liver lesions. However, the use of contrast agents in CE-CT scans imposes significant physical and economic burdens on patients in clinical practice. Recently, Generative Adversarial Networks (GANs)-based synthesis models offer an alternative approach that obtains CE-CT images from NC-CT images. However, poor coverage and mode collapse greatly limit their performance. Diffusion models (DMs)-based methods have demonstrated superior performance in natural image synthesis tasks. Nevertheless, our experiment shows that CE-CT images synthesized from DMs-based method exhibit higher overall quality but lower local quality. The quality of local areas, particularly those related to lesion areas, is crucial in medical image synthesis tasks. Hence, we propose a GANs-guided conditional diffusion model (GANs-CDM), combining the GANs and conditional diffusion model (CDM), to generate CECT images. In the proposed GANs-CDM, the GANs is to generate a preliminary CE-CT image, serving as conditional input for guiding the subsequent CDM to produce refined CE-CT images. Qualitative and quantitative evaluation on arterial and portal venous phase synthesis tasks demonstrates that our proposed GANs-CDM can significantly improve both the local and global quality of synthetic images. © 2024 IEEE.

关键词： Generative adversarial 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：