检索结果-内蒙古大学图书馆

7th International Conference on Inventive Computation Technologies, ICICT 2024

作者： Pawar, Poonam Ainapure, Bharati Vishwakarma University Faculty of Science And Technology Dept of Computer Engineering Pune India

ISBN: (纸本)9798350359299

Digital image quality is essential for a variety of applications in visual information processing, including medical diagnostics, defence, and many other applications. However, there are several challenges that must be overcome before an image can be used, and these often result in noise, distortion, and degradation. The field of image restoration is growing in significance as it focuses on repairing and restoring damaged images. With the increasing number of imaging equipment and the wide range of applications, image restoration techniques are becoming more and more important. Because of the challenges posed by low light levels, sensor noise, and other environmental limitations, more analysis on restoration strategies is needed. This study provides researchers, practitioners, and students with a clear understanding of image restoration techniques by providing an extensive overview of the developments in different fields. This research study offers a comprehensive analysis by analysing the development of these techniques, from traditional methodologies to the most recent deep learning approaches. Here we have proposed a novel architecture for image restoration, designed to facilitate seamless implementation and achieve high-quality results. © 2024 IEEE.

关键词： Image reconstruction

来源：评论

学校读者我要写书评

暂无评论

Topology-adaptive Bayesian optimization for deep ring echo state networks in speech emotion recognition

引用

Neural Computing and Applications 2025年第1期37卷 399-416页

作者： Soltani, Rebh Benmohamed, Emna Ltifi, Hela University of Sfax BP 1173 Sfax3038 Tunisia Department of computer science Onaizah college of Engineering and Information Technology Oniazah colleges Saudi Arabia Computer Science and Mathematics Department Faculty of Sciences and Techniques of Sidi Bouzid University of Kairouan Kairouan Tunisia

Speech emotion recognition is a difficult task that is gaining attention in a variety of domains, including psychology, human–computer interaction, and speech processing. To recognize speech emotions, machine learning techniques such as convolutional neural networks and long short-term memory networks have been successfully used. However, these models often fail to capture the long-term temporal dynamics of emotional speech. In this work, data augmentation was performed on all datasets. The generated result was used to extract the discriminant features. Then we apply a deep echo state network with a ring topology. The Bayesian optimization (BO) algorithm was applied to the spectral radius of the reservoir matrix. Our approach reveals term-weighted and unweighted accuracies of (75.45%, 74.79%), (86.89%, 86.14%), (98.99%, 97.62%), and (62%, 61.24%) on EMO-DB, SAVEE, RAVDESS, and TESS datasets. The findings demonstrate that by using BO, we are able to determine an optimal spectral radius. Therefore, this approach balances the minimal complexity of reservoir design and high emotion recognition accuracy. Notably, we demonstrate that a simple, deterministically constructed cycle reservoir within DRESN is comparable to other methodologies. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Bayesian networks

来源：评论

学校读者我要写书评

暂无评论

Fast Analysis of Characteristic Modes Using Hybrid Integral Equation Methods

Fast Analysis of Characteristic Modes Using Hybrid Integral ...

引用

2024 International Applied Computational Electromagnetics Society Symposium, ACES-China 2024

作者： He, Yusheng Shao, Hanru Ningbo University Faculty of Electrical Engineering and Computer Science Ningbo315211 China

ISBN: (纸本)9798350355581

To expedite the characteristic mode analysis (CMA) of electromagnetic target structures, this paper combines frequency- and material-independent reactions (FMIR) with characteristic mode analysis using the volume-surface integral equation (VSIE). Although the CMA based on VSIE is not affected by spurious modes when calculating targets containing dielectric, its matrix is high dimensional, resulting in slower calculations. FMIR only needs to store the calculation once for matrices that are independent of frequency and material parameters, which greatly reduces the repetitive operations when calculating multi-frequency impedance matrices. Numerical results show that combining FMIR with the CMA using VSIE can effectively improve the computational speed. © 2024 IEEE.

关键词： Matrix algebra

来源：评论

学校读者我要写书评

暂无评论

CLASS PROBABILITY MATCHING WITH CALIBRATED NETWORKS FOR LABEL SHIFT ADAPTATION 12

CLASS PROBABILITY MATCHING WITH CALIBRATED NETWORKS FOR LABE...

引用

12th International Conference on Learning Representations, ICLR 2024

作者： Wen, Hongwei Betken, Annika Hang, Hanyuan Faculty of Electrical Engineering Mathematics and Computer Science University of Twente Enschede Netherlands

We consider the domain adaptation problem in the context of label shift, where the label distributions between source and target domain differ, but the conditional distributions of features given the label are the same. To solve the label shift adaptation problem, we develop a novel matching framework named class probability matching (CPM). It is inspired by a new understanding of the source domain's class probability, as well as a specific relationship between class probability ratios and feature probability ratios between the source and target domains. CPM is able to maintain the same theoretical guarantees as the existing feature probability matching framework, while significantly improving the computational efficiency due to directly matching the probabilities of the label variable. Within the CPM framework, we propose an algorithm named class probability matching with calibrated networks (CPMCN) for target domain classification. From the theoretical perspective, we establish a generalization bound of the CPMCN method in order to explain the benefits of introducing calibrated networks. From the experimental perspective, real data comparisons show that CPMCN outperforms existing matching-based and EM-based algorithms. © 2024 12th International Conference on Learning Representations, ICLR 2024. All rights reserved.

关键词： Computational efficiency

来源：评论

学校读者我要写书评

暂无评论

Fake News Classification: Past, Current, and Future

引用

computers, Materials & Continua 2023年第11期77卷 2225-2249页

作者： Muhammad Usman Ghani Khan Abid Mehmood Mourad Elhadef Shehzad Ashraf Chaudhry Department of Computer Science University of Engineering and TechnologyLahore54890Pakistan Department of Computer Science&Information Technology Abu Dhabi UniversityAbu Dhabi59911United Arab Emirates Department of Software Engineering Faculty of Engineering and ArchitectureNisantasi UniversityIstanbulTurkey

The proliferation of deluding data such as fake news and phony audits on news web journals,online publications,and internet business apps has been aided by the availability of the web,cell phones,and social *** can quickly fabricate comments and news on social *** most difficult challenge is determining which news is real or ***,tracking down programmed techniques to recognize fake news online is *** an emphasis on false news,this study presents the evolution of artificial intelligence techniques for detecting spurious social media *** study shows past,current,and possible methods that can be used in the future for fake news *** different publicly available datasets containing political news are utilized for performing *** supervised learning algorithms are used,and their results show that conventional Machine Learning(ML)algorithms that were used in the past perform better on shorter text *** contrast,the currently used Recurrent Neural Network(RNN)and transformer-based algorithms perform better on longer ***,a brief comparison of all these techniques is provided,and it concluded that transformers have the potential to revolutionize Natural Language Processing(NLP)methods in the near future.

关键词： Supervised learning algorithms fake news classification online disinformation transformers recurrent neural network(RNN)disinformation transformers recurrent neural network(RNN)

来源：评论

学校读者我要写书评

暂无评论

Determining category metadata in open data portals - an approach based on Formal Concept Analysis 32

Determining category metadata in open data portals - an appr...

引用

32nd Telecommunications Forum, TELFOR 2024

作者： Gligorijevic, Milena Frtunic Bogdanovic, Milos Stoimenov, Leonid University of Niš Faculty of Electronic Engineering Department for Computer Science Niš Serbia

ISBN: (纸本)9798350391053

The efforts for data transparency and open government initiatives have resulted in a large amount of data being published on open data portals. These portals are organized to enhance published data accessibility by providing search mechanisms that utilize metadata such as category, tags, format, and publisher. However, due to the metadata incompleteness, such as missing category information, datasets are harder to find. Therefore, within this paper, we address the issue of missing category information by proposing an approach for categorizing datasets using keywords and their usage within categories. The presented approach is based on the Formal Concept Analysis method, which is used to create a hierarchical order of keywords grounded on their usage within categories on one open data portal. Further, we present the evaluation of the presented method using data from the Canadian open data portal. © 2024 IEEE.

关键词： categorization FCA formal concept analysis metadata enrichment open data open data portals

来源：评论

学校读者我要写书评

暂无评论

Transmission of Electromagnetic Waves Through a Clay Material 1

引用

32nd International Conference on Organization and Technology of Maintenance, OTO 2023

作者： Mandrić, Vanja Rupčić, Slavko Vinko, Davor Bilandžija, Domagoj Faculty of Electrical Engineering Computer Science and Information Technology Osijek Osijek Croatia

ISBN: (数字)9783031514944

ISBN: (纸本)9783031514937

This paper deals with the simulation calculation of transmission (S21) and reflection (S22) parameters in a material parametrically based on clay (brick). The electromagnetic parameters of the clay that are the subject of the study are: relative permittivity Εr, loss tangent tan δ and electrical conductivity σ, while the only structural parameter is the thickness d of the sample. The values of all considered parameters are within the range of data from available literature and values that could occur due to external causes (e. g. an increase in clay moisture leads to an increase in electrical conductivity). The simulation model was carried out with the software tool ANSYS HFSS and validated with a laboratory measurement model. The laboratory model consists of a circular waveguide with an inner diameter of 150 mm and a total length of 700 mm, which is excited by 52 mm long probes placed 50 mm from the last wall of the waveguide. The lower cut-off frequency of this measurement system is 1.173 GHz, so all results (simulation and measurement) are limited by this lower cut-off frequency of the waveguide. All simulation calculations were performed for three clay sample thicknesses: 18 mm, 36 mm and 54 mm. These calculations were made with the aim of determining brick composites that have high attenuation of electromagnetic EM waves (low transmission parameter and high reflection parameter) in order to reduce the strength of EM fields inside the object made of such bricks. The measured results were used to verify the simulation model. © 2024, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Circular waveguides

来源：评论

学校读者我要写书评

暂无评论

Traditional and parameter-efficient fine-Tuning of LLMs for sentiment analysis in the English and Serbian language 11

Traditional and parameter-efficient fine-Tuning of LLMs for ...

引用

11th International Conference on Electrical, Electronic and Computing engineering, IcETRAN 2024

作者： Dordevic, Nikola Stojkovic, Suzana University of Niš Faculty of Electronic Engineering Department of Computer Science Niš Serbia

ISBN: (纸本)9798350386998

Fine-Tuning of large language models is often demanding in terms of computational resources and memory. Consequently, there is a need to explore new methods that can effectively fine-Tune these models without compromising performance. In this paper, we compared traditional and parameter-efficient fine-Tuning of large language model XLM-RoBERTa, across both high-resourced languages like English and low-resourced languages such as Serbian. The model is fine-Tuned for sentiment analysis when it is conducted using text classification. The experiments have shown that parameter-efficient fine-Tuning can produce results that are on par with traditional fine-Tuning, both for the English and the Serbian language. Furthermore, parameter efficient fine-Tuning approach significantly reduces the computational time and memory requirements associated with model training, thereby offering a promising avenue for resource-efficient large language models fine-Tuning. In experiments with the English language, we achieved accuracy using the PEFT approach that was only 3% lower than the accuracy obtained through the traditional approach, while reducing training time by 35% and memory requirements by 42%. In experiments involving the Serbian language, we achieved accuracy using the PEFT approach that was only 1% lower than the accuracy obtained through the traditional approach, while reducing training time by 42% and memory requirements by 69%. © 2024 IEEE.

关键词： Modeling languages

来源：评论

学校读者我要写书评

暂无评论

Efficient Earth Observation Data Visualization Using Dual Orthogonal Equidistant Cylindrical Projection 11

Efficient Earth Observation Data Visualization Using Dual Or...

引用

11th International Conference on Electrical, Electronic and Computing engineering, IcETRAN 2024

作者： Dimitrijević, Aleksandar University of Niš Faculty of Electronic Engineering Department of Computer Science Niš Serbia

ISBN: (纸本)9798350386998

Earth observation (EO) data have seen a constant surge in volume, necessitating efficient storage, retrieval, and sharing mechanisms. Cartographic projections play a vital role in transforming spheroidal surface data into two-dimensional representations, albeit with distortions. The Dual Orthogonal Equidistant Cylindrical (DOEC) projection offers a balance between simplicity and distortion, making it a promising choice for EO data storage and visualization. This paper explores various methods for raster geodata visualization in DOEC projection, addressing challenges of partitioning, distortion, and visualization techniques. Through experimental comparisons, the paper demonstrates the efficiency of these methods, highlighting their applicability in EO data processing and visualization systems. © 2024 IEEE.

关键词： Rasterization

来源：评论

学校读者我要写书评

暂无评论

An Approach for Metadata Enrichment on Open Data Portals 11

An Approach for Metadata Enrichment on Open Data Portals

引用

11th International Conference on Electrical, Electronic and Computing engineering, IcETRAN 2024

作者： Gligorijević, Milena Frtunic Bogdanović, Milos Stoimenov, Leonid University of Niš Faculty of Electronic Engineering Department for Computer Science Niš Serbia

ISBN: (纸本)9798350386998

Open data initiatives have resulted in a large amount of data being published on open data portals. In order to make published data more accessible these portals provide search mechanisms based on metadata like category, tags, format, and publisher. However, due to the lack of metadata information, like category values, search results are incomplete, and data is harder to be found. Therefore, this paper presents an approach for metadata enrichment on open data portals. In particular, the presented approach is focused on suggesting categories one dataset should belong to, based on the combination of tags that is used for its description. This approach relies on the tags usage for describing datasets within categories and the Formal Concept Analysis method for creating a hierarchical order of tags within categories. The presented proposal has been evaluated using data from Ireland's open data portal. © 2024 IEEE.

关键词： FCA formal concept analysis metadata enrichment open data open data portals

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

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

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：