检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Dragar, Luka Peer, Peter Štruc, Vitomir Batagelj, Borut Faculty of Computer and Information Science University of Ljubljana Večna pot 113 LjubljanaSI-1000 Slovenia University of Ljubljana Faculty of Electrical Engineering Trzaška cesta 25 Ljubljana1000 Slovenia

In the era of rapid digitalization and artificial intelligence advancements, the development of DeepFake technology has posed significant security and privacy concerns. This paper presents an effective measure to assess the visual realism of DeepFake videos. We utilize an ensemble of two Convolutional Neural Network (CNN) models: Eva and ConvNext. These models have been trained on the DeepFake Game Competition (DFGC) 2022 dataset and aim to predict Mean Opinion Scores (MOS) from DeepFake videos based on features extracted from sequences of frames. Our method secured the third place in the recent DFGC on Visual Realism Assessment held in conjunction with the 2023 International Joint Conference on Biometrics (IJCB 2023). We provide an overview of the models, data preprocessing, and training procedures. We also report the performance of our models against the competition's baseline model and discuss the implications of our findings. Copyright © 2023, The Authors. All rights reserved.

关键词： Biometrics

来源：评论

学校读者我要写书评

暂无评论

Paramagnon heat capacity and anomalous thermopower in anisotropic magnetic systems: Understanding interlayer spin correlations in a magnetically disordered phase

引用

Physical Review B 2024年第5期109卷 054418-054418页

作者： Fatemeh Heydarinasab Morteza Jazandari Md Mobarak Hossain Polash Jahanfar Abouie Daryoosh Vashaee Department of Physics Faculty of Science University of Sistan and Baluchestan Zahedan 98167-45845 Iran Department of Physics Institute for Advanced Studies in Basic Sciences (IASBS) Zanjan 45137-66731 Iran Department of Materials Science and Engineering North Carolina State University Raleigh North Carolina 27606 USA Department of Electrical and Computer Engineering North Carolina State University Raleigh North Carolina 27606 USA

The interplay between entropy transport and charge carriers–paramagnon interaction in the Onsager linear system has been a subject of debate due to the limited theoretical and experimental understanding of paramagnon heat capacity. In this study, we investigate this interplay in an anisotropic layered magnetic system using cluster mean-field theory with spin quantum correlations. By examining spin correlation functions between different spins with various types of clustering, we derive the spin correlation function as a function of distance and temperature for the interlayer clusters both below and above the magnetic order phase transition. Our analysis reveals that paramagnons characterized by pronounced spin correlations among interlayer nearest-neighbor spins exhibit a nonzero heat capacity, providing valuable insights into the dynamics of entropy transport. The findings align with experimental observations, lending strong support to the validity of the paramagnon-drag thermopower concept. This study sheds light on the intricate dynamics and thermodynamic properties of paramagnons, advancing our understanding of entropy transport in complex systems.

关键词： Entropy Specific heat Spin dynamics Thermoelectric effects

来源：评论

学校读者我要写书评

暂无评论

RAINER: A Robust Ensemble Learning Grid Search-Tuned Framework for Rainfall Patterns Prediction

arXiv

引用

arXiv 2025年

作者： Li, Zhenqi Zhong, Junhao Wang, Hewei Xu, Jinfeng Li, Yijie You, Jinjiang Zhang, Jiayi Wu, Runzhi Dev, Soumyabrata The Fu Foundation School of Engineering and Applied Science Columbia University United States Biomedical Engineering Department Pratt School of Engineering Duke University United States School of Computer Science University College Dublin Ireland Department of Electrical and Electronic Engineering The University of Hong Kong Hong Kong Robotics Institute School of Computer Science Carnegie Mellon University United States Faculty of Science and Engineering University of Nottingham Ningbo China China College of Intelligence and Computing Tianjin University China The ADAPT SFI Research Centre

Rainfall prediction remains a persistent challenge due to the highly nonlinear and complex nature of meteorological data. Existing approaches lack systematic utilization of grid search for optimal hyperparameter tuning, relying instead on heuristic or manual selection, frequently resulting in sub-optimal results. Additionally, these methods rarely incorporate newly constructed meteorological features such as differences between temperature and humidity, to capture critical weather dynamics. Furthermore, there is a lack of systematic evaluation of ensemble learning techniques and limited exploration of diverse advanced models introduced in the past one or two years. To address these limitations, we propose a Robust ensemble LeArning grId search-tuNed framEwoRk (RAINER) for rainfall prediction. RAINER incorporates a comprehensive feature engineering pipeline, including outlier removal, imputation of missing values, feature reconstruction, and dimensionality reduction via Principal Component Analysis (PCA). The framework integrates novel meteorological features to capture dynamic weather patterns and systematically evaluates non-learning mathematical-based methods and a variety of machine learning models, from weak classifiers (e.g., KNN, LASSO, and Random Forest) to advanced neural networks such as Kolmogorov-Arnold Networks (KAN). By leveraging grid search for hyperparameter tuning and ensemble voting techniques, RAINER achieves state-of-the-art results within real-world dataset. Extensive experiments demonstrate the framework’s superior performance across diverse metrics, such as Accuracy, Precision, Recall, F1-score, and AUC, highlighting the pivotal role of feature engineering, grid search, and ensemble learning in rainfall prediction tasks. © 2025, CC BY.

关键词： Dimensionality reduction

来源：评论

学校读者我要写书评

暂无评论

An Enhanced Deep Super-Resolution Generative Adversarial Network Approach for Skin Disease Image Regeneration and Assessment

An Enhanced Deep Super-Resolution Generative Adversarial Net...

引用

Electronic engineering (ICEEM), International Conference on

作者： Ibrahim Abd El-Fattah Taha E. Taha Walid El-Shafai Department of Electronics and Electrical Communications Engineering Faculty of Electronic Engineering Menoufia University Menouf Egypt Department Electronics and Electrical Communications Engineering Alexandria Higher Institute of Engineering and Technology (AIET) Alexandria Egypt Computer Science Department Security Engineering Lab Prince Sultan University Riyadh Saudi Arabia

Skin cancer is a critical medical concern, posing significant challenges in accurate diagnosis. Algorithmic approaches have seen remarkable advancements across various occupations, including skin disease assessment. This study introduces a distinctive Improved Deep Super-Resolution Generative Adversarial Network (I_DSUR-GAN) methodology for regenerating low-resolution (LOR) skin disease images into super-resolution (SUR) format. Additionally, a modified SUR-dataset inspired by HAM10000 is presented for skin disease applications. The proposed approach incorporates a novel loss function design to provide supplementary information, facilitating the creation of high-quality SUR images. Experimental results demonstrate that our method outperforms existing approaches on the HAM10000 dataset. A comprehensive evaluation and employing sensitive metrics is conducted to assess the effectiveness, training periods, and memory requirements of the proposed framework. The outcomes reveal that the suggested model excels in restoring and identifying hue and texture compared to conventional and earlier models.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Cypher by Example: A Visual Query Language for Graph Databases

SSRN

引用

SSRN 2023年

作者： Rabuzin, Kornelije Šestak, Martina Faculty of Organization and Informatics University of Zagreb Pavlinska ulica 2 Varaždin42000 Croatia University of Maribor Faculty of Electrical Engineering and Computer Science Koroška cesta 46 Maribor2000 Slovenia

Considering the constantly increasing connectivity in terms of data exchanged between applications, devices etc., modern IT systems are in need of a data storage solution able to capture and handle data about connections and patterns formed between system participants (entities). As a potential solution for this challenge, graph database emerged in the previous decade. Since the standardization of a graph database query language is still an undergoing effort, any interested parties can interact with a graph database and explore graph data by using query languages such as Cypher or Gremlin, which are supported by modern Graph Database Management Systems (GDBMSs). Even though the syntax of these languages is well developed and documented, potential end-users with little knowledge of graph databases and the graph data model can still face a steep learning curve before being able to write queries on their own data stored in a form of a property graph. Consequently, this brings to a limited interest among the users to implement graph databases in their domains, as there is obviously a missing tool for maintaining the graph database in a visual manner. To address this gap, this paper introduces Cypher by Example, a visual graph query language consisting of a visual interface and a set of operations for interacting with the underlying graph database. The paper presents the basic elements of the proposed query language and demonstrates its effectiveness and completeness on a series of Cypher queries executed for a sample education domain. © 2023, The Authors. All rights reserved.

关键词： Query languages

来源：评论

学校读者我要写书评

暂无评论

Quantile-Based Maximum Likelihood Training for Outlier Detection

arXiv

引用

arXiv 2023年

作者： Taghikhah, Masoud Kumar, Nishant Šegvić, Siniša Eslami, Abouzar Gumhold, Stefan Faculty of Computer Science Technische Universität Dresden Dresden Germany Faculty of Electrical Engineering and Computing University of Zagreb Zagreb Croatia Translational Research Lab Carl Zeiss Meditec AG Munich Germany

Discriminative learning effectively predicts true object class for image classification. However, it often results in false positives for outliers, posing critical concerns in applications like autonomous driving and video surveillance systems. Previous attempts to address this challenge involved training image classifiers through contrastive learning using actual outlier data or synthesizing outliers for self-supervised learning. Furthermore, unsupervised generative modeling of inliers in pixel space has shown limited success for outlier detection. In this work, we introduce a quantile-based maximum likelihood objective for learning the inlier distribution to improve the outlier separation during inference. Our approach fits a normalizing flow to pre-trained discriminative features and detects the outliers according to the evaluated log-likelihood. The experimental evaluation demonstrates the effectiveness of our method as it surpasses the performance of the state-of-the-art unsupervised methods for outlier detection. The results are also competitive compared with a recent self-supervised approach for outlier detection. Our work allows to reduce dependency on well-sampled negative training data, which is especially important for domains like medical diagnostics or remote sensing. © 2023, CC BY.

关键词： Statistics

来源：评论

学校读者我要写书评

暂无评论

Power Quality 24-Hour Prediction Using Differential, Deep and Statistics Machine Learning Based on Weather Data in an Off-Grid

SSRN

引用

SSRN 2022年

作者： Zjavka, Ladislav Department of Computer Science Faculty of Electrical Engineering and Computer Science VŠB-Technical University of Ostrava 17. listopadu 15/2172 Ostrava Czech Republic

Prediction of Power Quality (PQ) on a daily basis is inevitable in planning the RE supply and scheduling the power load in smart off-grid autonomous systems. Various combinations of the attached power consumers lead to specific operating modes, related to the state of charge and load switching-time in different out-side conditions showing the system great variability. Complexity and irregularities require modelling with the use of artificial intelligence (AI) to represent the uncertainty in load and Renewable Energy (RE) ground fluctuations, which cannot be numerically solved with respect to unexpected states in detached systems. A recently designed and developed Differential Learning, using a novel neuro-computing approach, enables one to model high non-linear and indefinable chaotic physical systems. Its PQ-prediction results are presented in comparison to standard deep and probabilistic machine learning regression techniques. The model initialization allows operable day-ahead PQ-predictions in processing the last 24-hour series in one sequence. The proposed 2-level PQ optimization procedure evaluates initial load-scheduling plans in system management, based on PVP prediction. The efficiency and failure-free operation of the system are planned considering the first estimate of its day-ahead PVP supply and the secondary PQ verification of the system provided RE load utilization schemes, under various states. © 2022, The Authors. All rights reserved.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Prediction of Internet Broadband Connectivity Classification in West Java Indonesia using Machine Learning

Prediction of Internet Broadband Connectivity Classification...

引用

Artificial Intelligence Robotics, Signal and Image Processing (AIRoSIP), International Conference on

作者： Wahyu Haris Kusuma Atmaja Yaya Heryadi Ilvico Sonata Lukas Wayan Suparta Computer Science Department Binus Graduate Program - Computer Science Bina Nusantara University West Jakarta Indonesia Cognitive Engineering Research Group (CERG) Universitas Katolik Indonesia Atma Jaya Jakarta Indonesia Department of Electrical Engineering Faculty of Industrial Engineering Institut Teknologi Nasional Yogyakarta Yogyakarta Indonesia

ISBN: (数字)9798350327472

ISBN: (纸本)9798350327489

A location's Take-up Rate was significantly influenced by its Internet connectivity and availability. The purpose of this research is to answer concerns about internal Internet Service Provider issues that affect Internet connectivity in clusters of West Java, Indonesia. It also aims to evaluate the results of models that forecast how these internal factors will affect internet connectivity. According to the research findings, the FTTH Cluster has eight significant Internet predictor connectivity out of 18 elements. Activation delay and Administration Documentation of Activation with Connectivity were the two major parameters modeled out of the 18 components. Artificial Neural Network (ANN), Support Vector Machine (SVM), and Decision Tree (DT) models were used to predict Internet Broadband Connectivity Classification based on the two most important variables. According to the study, ANN is the best model, with 99% accuracy and 99.6% precision when compared to SVM and DT. Furthermore, it exceeds previous studies that used the ANN model and achieved 97.92% accuracy.

关键词： Support vector machines Java Accuracy Web and internet services Artificial neural networks Predictive models Internet Broadband communication Optical fiber subscriber loops Robots

来源：评论

学校读者我要写书评

暂无评论

Efficient and High-quality Recommendations via Momentum-incorporated Parallel Stochastic Gradient Descent-Based Learning

引用

IEEE/CAA Journal of Automatica Sinica 2021年第2期8卷 402-411页

作者： Xin Luo Wen Qin Ani Dong Khaled Sedraoui MengChu Zhou the School of Computer Science and Technology Dongguan University of TechnologyDongguan 523808 Hengrui(Chongqing)Artificial Intelligence Research Center Department of Big Data Analyses TechniquesCloudwalkChongqing 401331China the School of Computer Science and Technology Chongqing University of Posts and TelecommunicationsChongqing 400065 Chongqing Engineering Research Center of Big Data Application for Smart Cities Chongqing Institute of Green and Intelligent TechnologyChinese Academy of SciencesChongqing 400714China the Department of Computer and Information Science City College of Dongguan University of TechnologyDongguan 523419China the Department of Electrical and Computer Engineering Faculty of Engineeringand Center of Research Excellence in Renewable Energy and Power SystemsKing Abdulaziz UniversityJeddah 21481Saudi Arabia the Department of Electrical and Computer Engineering New Jersey Institute of TechnologyNewarkNJ 07102 USA the Center of Research Excellence in Renewable Energy and Power Systems King Abdulaziz UniversityJeddah 21481Saudi Arabia

A recommender system(RS)relying on latent factor analysis usually adopts stochastic gradient descent(SGD)as its learning ***,owing to its serial mechanism,an SGD algorithm suffers from low efficiency and scalability when handling large-scale industrial *** at addressing this issue,this study proposes a momentum-incorporated parallel stochastic gradient descent(MPSGD)algorithm,whose main idea is two-fold:a)implementing parallelization via a novel datasplitting strategy,and b)accelerating convergence rate by integrating momentum effects into its training *** it,an MPSGD-based latent factor(MLF)model is achieved,which is capable of performing efficient and high-quality *** results on four high-dimensional and sparse matrices generated by industrial RS indicate that owing to an MPSGD algorithm,an MLF model outperforms the existing state-of-the-art ones in both computational efficiency and scalability.

关键词： Big data industrial application industrial data latent factor analysis machine learning parallel algorithm recommender system(RS) stochastic gradient descent(SGD)

来源：评论

学校读者我要写书评

暂无评论

New In-Air Signature Datasets

New In-Air Signature Datasets

引用

International Symposium on Networks, computers and Communications (ISNCC)

作者： AbdulAzeez R. Alobaidi Thameur Dhieb Zeina N. Nuimi Tarek M. Hamdani Ali Wali Adel M. Alimi REGIM-Lab.: REsearch Groups in Intelligent Machines Enis University of Sfax Tunisia Department of Computer Techniques Engineering Faculty of Information Technology Imam Ja'afar Al-Sadiq University Iraq Higher Institute of Business Administration University of Gafsa Tunisia Department of Electrical Engineering College of Engineering University of Misan Iraq Higher Institute of Computer Science Mahdia University of Monastir Tunisia Higher Institute of Computer Science and Multimedia of Sfax University of Sfax Tunisia Department of Electrical and Electronic Engineering Science Faculty of Engineering and the Built Environment University of Johannesburg South Africa

Compared to traditional biometric systems, in-air signatures are considered more robust and secure than classical pen paper. A few datasets capturing in-air signatures have been introduced, utilizing various devices such as the Leap Motion and the Microsoft Kinect sensor camera. However, these devices are not exempt from shortcomings and exhibit certain limitations. The expenses associated with their implementation and the requirement for technical proficiency in operating them present notable challenges for in-air signature analysis. Additionally, users may encounter difficulties in adapting their finger movements to fit within the device's limited field of view, particularly if they lack familiarity with these devices. To address these concerns, this paper proposes the creation of three in-air signature datasets using solely the camera of a laptop or a smartphone, eliminating the need for any additional specialized equipment. Our datasets were collected in three ways. The first is the In-Air Signature dataset (IAS dataset) and the second is the In-Air Signature dataset using a transparent Glass Plate (IASGP dataset) while the third is the In-Air Signature dataset using Smart Phone (IASSP dataset). Forty volunteers participated in the construction of these datasets. Their ages ranged from 21 to 40 years. Each volunteer signs in the air five signatures and imitates five signatures of five other volunteers. Our in-air signatures datasets are publicly available and can be used for various research tasks like in-air signature verification and identification.

关键词：

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：