检索结果-内蒙古大学图书馆

Hierarchical Attention Mechanism for Domain-invariant Spoken Language Identification via Inter-domain Alignment and Intra-domain Language Discrimination

Hierarchical Attention Mechanism for Domain-invariant Spoken...

引用

engineering Informatics, IEEE

作者： Urvashi Goswami H Muralikrishna Sujeet Kumar A. D. Dileep Veena Thenkanidiyoor School of Computing and Electrical Engineering Indian Institute of Technology Mandi Mandi India Department of Electronics and Communication Engineering Manipal Academy of Higher Education Manipal Institute of Technology Manipal India Department of Computer Science and Engineering Indian Institute of Technology Dharwad Dharwad India Department of Computer Science and Engineering National Institute of Technology Goa Goa India

ISBN: (数字)9798331505776

ISBN: (纸本)9798331505783

Modern methods for spoken language identification (LID) have demonstrated promising results when trained on large datasets. However, their effectiveness is considerably impacted by the discrepancies between the training and testing distributions. Unsupervised domain adaptation (UDA) addresses the domain discrepancies by aligning feature distributions across source and target domains without using labeled target data, thereby ensuring the model performs consistently across both domains. This alignment, however, may come at the cost of reduced model discriminability. To enhance performance under domainmismatched conditions, a LID system should learn domaininvariant representations of speech while maintaining language discriminability at every level, from initial segment-level representations to final utterance-level representations. To address this issue, in this paper, we propose a hierarchical attention network that simultaneously enhances model domain invariance through inter-domain alignment and improves model’s language discriminability using intra-domain contrastive discrimination. The performance on the target domain shows that the proposed method exceeds the effectiveness of the top-performing existing approaches.

关键词： Training Adaptation models Computational modeling System performance Refining Speech enhancement Transformers Data models Optimization Testing

来源：评论

学校读者我要写书评

暂无评论

On-Demand Frequency Reconfigurable Flexible Antenna for 5Gsub-6-GHz and ISM Band Applications 6th

On-Demand Frequency Reconfigurable Flexible Antenna for 5Gsu...

引用

6th International Conference on Wireless Technologies, Embedded and Intelligent Systems, WITS 2020

作者： Hussain, Musa Rizvi, Syed Naheel Raza Awan, Wahaj Abbas Husain, Niamat Halima Hameed, Ahsan Department of Electrical Engineering Bahria University Islamabad Pakistan Department of Electronics Engineering University of Engineering and Technology Taxila Taxila Punjab Pakistan Department of Integrated IT Engineering Seoul National University of Science and Technology Seoul Korea Republic of Department of Computer and Communication Engineering Chungbuk National University Cheongju Korea Republic of Department of Telecommunication Engineering Baluchistan University of Information Technology Engineering and Management Sciences Quetta Pakistan Research Institute of Microwave and Millimeter-Wave Studies National University of Science and Technology Islamabad Pakistan

ISBN: (纸本)9789813368927

This paper presents the design and characterization of on-demand frequency reconfigurable antenna for 5Gsub-6-GHz and ISM-band applications. The antenna comprises of compact overall size having dimensions of 0.23λ0 × 0.2λ0 × 0.002λ0. The octagonal patch was loaded with stub to enhance bandwidth along with improvement in return loss, afterword a slot and diode was utilized to achieve frequency reconfigurability. The numerical analysis of the proposed antenna was done using Higher Frequency Structural Simulator (HFSS). Furthermore, to demonstrate the potential of the proposed work it is compared with state of the art works for similar applications. © 2022, Springer Nature Singapore Pte Ltd.

关键词： Microwave antennas

来源：评论

学校读者我要写书评

暂无评论

Optimal scheduling of ESS in distribution system with multiple DG 13

Optimal scheduling of ESS in distribution system with multip...

引用

13th International Conference on electrical and electronics engineering, ELECO 2021

作者： Tunçel, Suleyman Seker, Ayse Aybike Oskouei, Morteza Zare Gozel, Tuba Mohammadi-Ivatloo, Behnam Abapour, Mehdi Gebze Technical University Department of Electronics Engineering Kocaeli41400 Turkey Department of Electrical and Electronics Engineering Iskenderun Technical University Ratay 31200 Turkey Smart Energy Systems Laboratory University of Tabriz Faculty of Electrical and Computer Engineering Tabriz Iran

ISBN: (纸本)9786050114379

Increasing tendency to highly penetrate distribution networks with renewables due to environmental concerns, has lead distribution network operators to establish energy storage systems (ESS) for obtaining required flexibility. In this paper, the role of a widely used ESS, namely, compressed-air energy storage (CAES) is evaluated considering hosting distributed generation capacity of network. A deterministic scheduling strategy is proposed to minimize cost of operation and maintenance of network in presence of PV and wind generation as renewable sources. By utilization of CAES, renewable energy curtailment has been reduced while total electricity cost injected from the grid is minimized. Optimal charge and discharge scheduling of CAES has allowed excess renewable energy to be stored and then injected to the network on peak hours in order to reduce cost by decreasing power grid dependency on peak hours. © 2021 Chamber of Turkish electrical Engineers.

关键词： Compressed air

来源：评论

学校读者我要写书评

暂无评论

RSSI and Machine Learning-Based Indoor Localization Systems for Smart Cities

RSSI and Machine Learning-Based Indoor Localization Systems ...

引用

Artificial Intelligence (SLAAI-ICAI), SLAAI International Conference on

作者： R M M R Rathnayake M W P Maduranga Maheshi B. Dissanayake Department of Computer Engineering General Sir John Kotelawala Defence University Ratmalana Sri Lanka Department of Electrical and Electronics Engineering University of Peradeniya Kandy Sri Lanka

Widespread use of the Internet of Things (IoT) is being made in a number of application areas to save costs and improve performance. These types of applications also include services that are dependent on localization. When a sensor node is localized, an IoT network may use that information to identify it. A Received Signal Strength Indicator (RSSI), which gauges the intensity of the signal received from the sensor nodes, is typically used in conjunction with supervised Machine Learning (ML) algorithms in localization methods. In this study, a unique indoor and outdoor localization and navigation technique is presented that makes use of LoRA nodes and a highly accurate machine learning model. The suggested method has cutting-edge features and shows considerable gains in accuracy and dependability. These discoveries offer potential answers for practical applications and boost localization technology.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Bioinspired Sandwich Structure in Composite Panels

引用

Advances in Materials Science and engineering 2023年第1期2023卷

作者： Sampathkumar, Deepak Mohankumar, Ashokkumar Teekaraman, Yuvaraja Kuppusamy, Ramya Radhakrishnan, Arun Department of Mechanical and Automation Engineering Agni College of Technology Talambur Tamil Nadu Chennai600130 India Department of Mechanical Engineering Government College of Engineering Tamil Nadu Bargur635104 India Jahra Kuwait Department of Electrical and Electronics Engineering Sri Sairam College of Engineering Bangalore562106 India Faculty of Electrical & Computer Engineering Jimma Institute of Technology Jimma University Jimma Ethiopia

The phenomenon of separation into constituent layers connecting the core and laminate of a composite sandwich complex is a vital complication that leads to early failure of such material. The direction of the sandwich construction's exfoliation rigidity is increased between interlaminar low fiber augmentation. The bioinspired technique of hybrid material layers was used on an aluminium face sheet with an interlayer composition of PET foam core and glass fabric of a material that appears to have greater potential as a flimsy substitute for materials currently used in automotive, aeronautical, and marine applications. This examination seeks to develop the making of such material along the retardation in fibre supplements. Fibre bridging has been recognized as an important appliance in the progress of this operating procedure. Consequently, this method points to promoting the event of fibre bridging by differing aggregates, including the mass and extent of augmented fibres and the quantity of epoxy resin applied. A few advancements were made to the production methods, and though the outcomes for the resisting ability of specimens were found to be indecisive, it was found that the layer separation hardness had even improved. This was confirmed through the operation of scanning electron microscopy and also predicted the mechanically peeled material surfaces which identified the adhesive strength variations with respect to the face sheet surface modified with the sand blasting process. The analysis also revealed the need for further research into optimizing the attachment between aluminium sheet and pet foam and glass fabric based hybrid sandwich panels. © 2023 Deepak Sampathkumar et al.

关键词： Fibers

来源：评论

学校读者我要写书评

暂无评论

Design of A Hybrid Renewable Energy System with IoT Monitoring and Battery Management

Design of A Hybrid Renewable Energy System with IoT Monitori...

引用

International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (HNICEM)

作者： Caile Sofia Cabading John Francis Natividad Ronald Vincent Santiago Department of Electrical Electronics and Computer Engineering Mapùa University Manila Philippines

Two seasons that the Philippine archipelago experiences are: the wet and dry seasons. The country's mean annual rainfall ranges from 965 to 4,064 millimeters, according to PAGASA. The Philippine archipelago also takes advantage of abundant sunlight. From the data gathered from the NREL, the Philippines' average solar irradiance measured in watts per square meter is around 128-203, which provides a 4.5-5.5-kilowatt-hour per square meter daily power generating capacity. The research focuses on designing a Hybrid Renewable Energy System Harvester with IoT Monitoring and Battery Management. Specifically, the research aims (i) to utilize the use of solar energy and the droplets from the rain to generate electricity; (ii) to create a Dual-Input Single Output (DISO) system that will incorporate the two input hybrid renewable sources (Solar and Piezo), and to output a single DC source; (iii) to create an IoT based system for monitoring and acquiring data; and (iv) to create a battery management system to monitor and protect the battery from overcharging. Three tests were done for the prototype: controlled and uncontrolled trials for the Piezoelectric Rain Energy harvester and voltage regulation for the Hybrid System. The highest voltage output measured from the piezo is 3.29V. The data from the voltage regulation shows that the voltage of the battery is gradually increasing, which means that the Hybrid System supplies voltage to the battery.

关键词： Meters Renewable energy sources Voltage measurement Rain Battery management systems Solar energy Battery charge measurement

来源：评论

学校读者我要写书评

暂无评论

Enhancing Remote Sensing Image Classification Using Arithmetic Optimization Algorithm with Deep Learning Approach

Enhancing Remote Sensing Image Classification Using Arithmet...

引用

International Conference on Advances in Information Technology (ICAIT)

作者： R. Shanthi Balasubramaniam D R. Venkatesh C. Sankari V. Manimaran Shaik Hedayath Basha Department of computer applications Faculty of science and humanities SRM Institute of science and Technology Chengalpattu Tamilnadu Department of Biomedical Engineering VelTech Rangarajan Dr Sagunthala R & D Institute of Science and Technology Chennai Computer Science and Engineering Ramco Institute of Technology Rajapalayam Electrical and Electronics Engineering Chennai Institute of Technology Department of Computer Science and Engineering Kongu Engineering College Erode Tamilnadu Department of ECE R.M.K. College of Engineering and Technology

ISBN: (数字)9798350383867

ISBN: (纸本)9798350383874

The classification of Remote sensing image (RSI)is a vital task in field of geospatial analysis where aerial or satellite images are classified into predefined classes or land cover sorts. It plays a main part in diverse fields namely agriculture, disaster management, environmental surveillance, urban planning, and natural resource management. With the machine learning (ML) and deep learning (DL) models, these images examined and categorized depending on their spatial, spectral and temporal features. The exact detection of remote sensing images simplifies decision-making procedures, aiding governments, environmental agencies, and companies to make knowledgeable decisions and policies. It authorizes experts to observe variations in lands, discover land use trends, and evaluate environmental effects and making it a vital device in addressing present difficult challenges connected to weather change, urban development, and natural resource protection. This study presents an enhanced remote sensing image classification utilizing arithmetic optimization algorithm with DL (ERSC-AODL) Approach. The ERSC-AODL technique concentrates on the accurate recognition of the RSI into dissimilar classes. It begins with the use of an EfficientNet feature extractor, which derives complex spectral and spatial data, and acts as a base of proper representation of the images. Moreover, the recognition process takes place by a bidirectional long short-term memory (BiLSTM). Finally, the hyperparameter optimization procedure is performed by the ERSC-AODL technique. To portray the proficient classification outputs of the ERSC-AODL technique, a range of experiments were utilized. The investigational outputs stated that the ERSC-AODL technique attains improved accomplishment over other models.

关键词： Natural resources Recurrent neural networks Feature extraction Spatial databases Classification algorithms Data mining Remote sensing Optimization Long short term memory Image classification

来源：评论

学校读者我要写书评

暂无评论

Solving Differential Equations with Advanced Algorithmic Techniques

Solving Differential Equations with Advanced Algorithmic Tec...

引用

Intelligent Control, Computing and Communications (IC3), International Conference on

作者： Alok Jain Taqi Al-Rubaye R S Sengar Bhukya Balakrishna Narendra Kumar Rajireddy Soujanya Gopal K Lovely Professional University Phagwara India Department of Medical Laboratory Technology College of Medical Technologies The Islamic University Najaf Iraq Department of Electrical & Electronics Engineering GL Bajaj Group of Institutions Mathura Department of Computer Science and Engineering MLR Institute of Technology Hyderabad Telangana India Department of Electrical Electronics & Communication Engineering Galgotias university Gerater Noida Uttar Pradesh Department of CSE GRIET Bachupally Hyderabad Telangana India New Horizon College of Engineering Bangalore India

ISBN: (数字)9798331527495

ISBN: (纸本)9798331527501

One big change that has happened in differential equations is the use of new computer methods to solve hard problems. The growth of differential equations has been helped by this effect. If these strategies hadn't been used during the process, there would have been no change at all. In this abstract, neural networks and adaptive differential equation solvers are used together. This is one of the new ideas being looked into within the limits of this abstract. You can't use any other way to solve this type of dynamic net ODE. There isn't any other good answer. For people who want to understand how the process works, this outline also includes a description of the method. The goal of this study is to show that the suggested plan is better than other methods by looking at six different ways. The point of this study is to show why the suggested approach is better. The similarity that was just shown is one example of how speed has gotten better.

关键词： Technological innovation Adaptive systems Heuristic algorithms Neural networks Process control Differential equations Computational efficiency Dynamical systems Standards Intelligent control

来源：评论

学校读者我要写书评

暂无评论

A Survey of Anomaly Detection in In-Vehicle Networks

arXiv

引用

arXiv 2024年

作者： Özdemir, Övgü Tugberk Işyapar, M. Karagöz, Pınar Schmidt, Klaus Werner Demir, Demet Alpay Karagöz, N. R&D Department Proven Technologies Inc Ankara Turkey Department of Computer Engineering Middle East Technical University Ankara Turkey Department of Electrical-Electronics Engineering Middle East Technical University Ankara Turkey

Modern vehicles are equipped with Electronic Control Units (ECU) that are used for controlling important vehicle functions including safety-critical operations. ECUs exchange information via in-vehicle communication buses, of which the Controller Area Network (CAN bus) is by far the most widespread representative. Problems that may occur in the vehicle's physical parts or malicious attacks may cause anomalies in the CAN traffic, impairing the correct vehicle operation. Therefore, the detection of such anomalies is vital for vehicle safety. This paper reviews the research on anomaly detection for in-vehicle networks, more specifically for the CAN bus. Our main focus is the evaluation of methods used for CAN bus anomaly detection together with the datasets used in such analysis. To provide the reader with a more comprehensive understanding of the subject, we first give a brief review of related studies on time series-based anomaly detection. Then, we conduct an extensive survey of recent deep learning-based techniques as well as conventional techniques for CAN bus anomaly detection. Our comprehensive analysis delves into anomaly detection algorithms employed in in-vehicle networks, specifically focusing on their learning paradigms, inherent strengths, and weaknesses, as well as their efficacy when applied to CAN bus datasets. Lastly, we highlight challenges and open research problems in CAN bus anomaly detection. © 2024, CC BY.

关键词： Buses

来源：评论

学校读者我要写书评

暂无评论

Ambiguous Fuzzy Einstein Ordered Averaging Operator: Application to the classification of Power Generation methods 4

Ambiguous Fuzzy Einstein Ordered Averaging Operator: Applica...

引用

4th IEEE International Conference on Advance Computing and Innovative Technologies in engineering, ICACITE 2024

作者： Malik, Shilpam Joshi, B.P. Bisht, Vandana Kharayat, Pushpendra Singh Goyal, Shweta Maharaja Surajmal Institute of Technology Dept. of Electrical & Electronics Engg New Delhi110058 India Department of Mathematics Graphic Era Hill University Bhimtal263132 India Deptt. of Computer Science & Enggineering Graphic Era Hill University Bhimtal263132 India Amrapali Group of Institutes Department of Mathematics Siksha Nagar Haldwani India Graphic Era Deemed to Be University Department of Electrical Engineering Dehradun248002 India

ISBN: (纸本)9798350360165

The classification of power generation methods (PGMs) can be treated as multi-criteria-decision-making-procedure (MCDMP) as it is affected by many. The information generations are always big task to stockholders due to the incomplete information as it is very hard to extract the exact information. Most of the things are qualitative and quantitative. Therefore, to study systems under uncertainty fuzzy set (FS) and intuitionistic FS (IFS) are presented as a breakthrough. The Ambiguous Fuzzy set theory (AFST) is presented as the generalization of FS consisting of four types of membership degrees. The AFST is powerful than the FS and IFS due to four different degrees of memberships. This study presents the Einstein Ordered Weighted Averaging (OWA) Operator (EOWAO) based on the Einstein Sum (ES) and OWA under the domain of AFST, which is termed as 'AFSeow'. Then, to apply the AFS eow operator to real life problems, a framework is presented to include it in the information fusion of decision making (DM) process. Moreover, the classification of Power Generation methods (PGMs) is carried out using the proposed operator and proposed framework to show its applicability. The obtained results and comparative analysis show the feasibility and validity of the AFSEow operator. © 2024 IEEE.

关键词： Fuzzy set theory

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：