检索结果-内蒙古大学图书馆

Multimedia Evaluation Benchmark Workshop 2020, MediaEval 2020

作者： Wang, Xinsheng Zhu, Jihua Scharenborg, Odette School of Software Engineering Xi'an Jiaotong University Xi'an China Multimedia Computing Group Delft University of Technology Delft Netherlands

No-audio Multimodal Speech Detection is one of the tasks in Media- Eval 2020, with the goal to automatically detect whether someone is speaking in social interaction on the basis of body movement signals. In this paper, a multimodal fusion method, combining signals obtained by an overhead camera and a wearable accelerometer, was proposed to determine whether someone was speaking. The proposed system directly takes the accelerometer signals as input, while using a pre-trained 3D convolutional network to extract the video features that work as input. Experiments on the No-audio Multimodal Speech Detection task show that our method outperforms all submissions of previous years. © 2020 Copyright 2020 for this paper by its authors. All Rights Reserved.

关键词： Accelerometers

来源：评论

学校读者我要写书评

暂无评论

A Simple Confidence-Supervised Model for High-Resolution Defect Recognition

A Simple Confidence-Supervised Model for High-Resolution Def...

引用

Robotics, Control and Automation (ICRCA), International Conference on

作者： Hui-Yue Yang Hui Chen Fan Zhang Xiang Tian Luning Zhu Guiguang Ding Research Center for Information Science and Technology School of Software Tsinghua University and Beijing National Tsinghua University Zhuoxi Institute of Brain and Intelligence Beijing China Beijing National Research Center for Information Science and Technology Tsinghua University Beijing China Biomedical Engineering and Instrument Science Zhejiang University Zhejiang China Group Corporation The 21th Research Institute of China Electronics Technology Shanghai China

ISBN: (数字)9798350344721

ISBN: (纸本)9798350344738

This paper addresses the critical role of chip manufacturing in advancing the high-tech industry. It emphasizes the need for quality control in the delicate and complex chip manufacturing process, particularly in the detection of surface defects on chip package images. Such detection not only ensures chip quality, but also increases production efficiency and reduces costs. This study explores the application of deep learning for chip defect detection, focusing on high-resolution images with small defect areas. The proposed classification algorithm, which is based on model confidence, consists of two key components: Single-Scale Defect Recognition (SSDR) and Cross-Scale Model Fusion (CSMF). First, the chip image is pre-processed and partitioned into multi-scale clippings. Then, a ResNet-18 model is trained on these different scale clippings. Confidence thresholds for positive and negative cases are set to improve the accuracy. The final defect classification decision is made by considering the results of the multi-scale clippings together. Experimental results show that both methods significantly improve the classification accuracy compared to the direct evaluation of the clippings. In particular, CSMF outperforms SSDR by 16.78%.

关键词： Deep learning Industries Accuracy Manufacturing processes Production Quality control Manufacturing

来源：评论

学校读者我要写书评

暂无评论

Large Scale Foundation Models for Intelligent Manufacturing Applications: A Survey

arXiv

引用

arXiv 2023年

作者： Zhang, Haotian Dereck, Semujju Stuart Wang, Zhicheng Lv, Xianwei Xu, Kang Wu, Liang Jia, Ye Wu, Jing Long, Zhuo Liang, Wensheng Ma, X.G. Zhuang, Ruiyan The College of Information Science and Engineering Northeastern University Shenyang110819 China The Foshan Graduate School Northeastern University Foshan528311 China The School of mechanical engineering and automation South China University of Technology Shenyang110819 China The School of Software Engineering South China University of Technology Guangzhou510006 China The Departement of Computer and Information Science Qinghai University of Science and Technology Xining China The Enterprise AI Midea Group Foshan528311 China

'Although the applications of artificial intelligence especially deep learning had greatly improved various aspects of intelligent manufacturing, they still face challenges for wide employment due to the poor generalization ability, difficulties to establish high-quality training datasets, and unsatisfactory performance of deep learning methods. The emergence of large scale foundational models(LSFMs) had triggered a wave in the field of artificial intelligence, shifting deep learning models from single-task, single-modal, limited data patterns to a paradigm encompassing diverse tasks, multimodal, and pre-training on massive datasets. Although LSFMs had demonstrated powerful generalization capabilities, automatic high-quality training dataset generation and superior performance across various domains, applications of LSFMs on intelligent manufacturing were still in their nascent stage. A systematic overview of this topic was lacking, especially regarding which challenges of deep learning can be addressed by LSFMs and how these challenges can be systematically tackled. To fill this gap, this paper systematically expounded current statue of LSFMs and their advantages in the context of intelligent manufacturing. and compared comprehensively with the challenges faced by current deep learning models in various intelligent manufacturing applications. We also outlined the roadmaps for utilizing LSFMs to address these challenges. Finally, case studies of applications of LSFMs in real-world intelligent manufacturing scenarios were presented to illustrate how LSFMs could help industries, improve their efficiency. Copyright © 2023, The Authors. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Dynamic Economic Dispatch Integrating Photovoltaic and Wind Generation Using GAMS Optimization

Dynamic Economic Dispatch Integrating Photovoltaic and Wind ...

引用

Wireless and Optical Technologies (GCWOT), 2020 Global Conference on

作者： Summaya Qasim Muhammad Mansoor Ashraf Syeda Sadaf Zaidi Muhammad Waseem Fabiano Pallonetto Zain Anwar Ali Electrical Engineering Department University of Engineering and Technology Taxila Pakistan Department of Electronics Engineering International Renewable and Energy Systems Integration Research Group (IRESI) Maynooth University Co. Kildare Ireland Lero - Science Foundation Ireland Research Centre for Software School of Business Innovation Value Institute (IVI) Maynooth University Co. Kildare Ireland Maynooth International Engineering College Maynooth University Co. Kildare Ireland

ISBN: (数字)9798331534271

ISBN: (纸本)9798331534288

This research presents a strategy to optimize per-formance of the conventional power grid systems incorporating photovoltaic and wind energy sources using Dynamic Economic Load Dispatch (ELD). The proposed model tackles the uncertain and variable characteristics of wind and solar power outputs by accounting for disparities between observed and forecasted power outputs leveraging General Algebraic Modeling System (GAMS)-based optimization. The approach seeks to reduce the overall operational expenses of generators while effectively distributing power generation to committed units. This methodology is applied to a power system incorporating the solar and wind plant linked to 3 and 6 generating units, respectively. The outcomes of simulations carried out in GAMS substantiate the efficacy of the proposed approach in terms of least fuel cost. GAMS is a commercial package deployed to optimize the ELD problem effectively with an easy-to-use interface which efficiently finds the optimal results using non-linear programming solver.

关键词： Photovoltaic systems Wireless communication Costs Wind energy Power system dynamics Predictive models Fuels Wind forecasting Optimization Load modeling

来源：评论

学校读者我要写书评

暂无评论

Application of K-Nearest Neighbors Algorithms for Void Classification in Composite Oriented Strand Board 9

Application of K-Nearest Neighbors Algorithms for Void Class...

引用

9th Annual Composites and Advanced Materials Expo, CAMX 2023

作者： Hu, Wenyue Eftekhari, Yaser Callander, Sam Wang, Xiaoxing Bowland, Christopher C. Nguyen, Frank McCaslin, Jeremy Schaal, Christoph Gu, Grace X. Li, Carina Jin, Bo Advanced Composites Design Lab University of Southern California United States Volume Graphics Software Fiber Composite Material Analysis Hexagon CharlotteNC United States MATLAB MathWorks Japan Chemical Sciences Division Carbon and Composites Group Oak Ridge National Lab United States The Gill Corporation El MonteCA United States ANSYS United States Department of Mechanical Engineering California State University Northridge United States Department of Mechanical Engineering University of California Berkeley United States Department of Aeronautics and Astronautics Massachusetts Institute of Technology United States

ISBN: (纸本)9781934551448

Composite Oriented Strand Board (COSB) is an aerospace-grade material that is manufactured using laminated, unidirectional carbon fiber-epoxy prepreg strands. Its manufacturing methodology allows fine-tuning, producing controllable thickness, flatness, and microstructure with quality assurance. To comprehend the microstructure of COSB from X-ray computed tomography (XCT) scan data, it is critical to accurately assess and quantify void content in the post-analysis. The conventional methods of post-analysis used for X-ray micro-computed tomography (micro-CT) have become inadequate due to their time-consuming nature and oftentimes imprecise measurements. In this paper, we present a new approach for enhancing void classification accuracy by utilizing the K-Nearest Neighbors Classifiers (KNN) with the assistance of convolutional kernels. KNN training based on two labels of greyscale thresholding images achieved a 1% error rate, a significant improvement over the three deep learning algorithms (Fully Convolutional Neural Network, U-net, SegNet) in our previous study. When classifying five different void labels, the multi-classifier KNN algorithm with convolutional kernels yields around 2% error rate. Encouraged by these results, we propose using convolutional neural networks (CNNs) to make more complex reasoning and decisions when classifying voids, improving the accuracy of void characterization. Copyright © 2023. Used by CAMX - The Composites and Advanced Materials Expo with permission.

关键词： Microstructure

来源：评论

学校读者我要写书评

暂无评论

Research on Train Integrity Monitoring Using Multi-constellation Measurements 12th

Research on Train Integrity Monitoring Using Multi-constella...

引用

12th China Satellite Navigation Conference, CSNC 2021

作者： Li, Jialei Liu, Yongqiang Jiang, Wei Cai, Baigen Wang, Jian School of Electronic and Engineering Beijing Jiaotong University Beijing100044 China Beijing Engineering Research Center of EMC and GNSS Technology for Rail Transportation State Key Laboratory of Rail Traffic Control and Safety School of Electronic and Engineering Beijing Jiaotong University Beijing100044 China General Technology Branch of CETC Northwest Group Software Products Department Xian710068 China

ISBN: (纸本)9789811631375

Train integrity monitoring is the key to ensure the safe operation of trains. Traditional train integrity monitoring is based on ground equipment. However, with the development of satellite positioning technology, and in order to meet the needs of the next generation train control system (NGTC), the demand for ground equipment should be minimized. The application of satellite positioning technology in train positioning is more and more. Its principle is to calculate the length of the moving-baseline between the head and end antennas of the train by satellite positioning, and monitor the integrity status of the train according to the length change. The multi-constellation positioning has more visible satellites and better geometric distribution of satellites compared with the single constellation system. Especially, the advantages of the multi-constellation system are more prominent in the limited satellite signals scenes. Therefore, this paper proposes a train integrity monitoring method based on multi-constellation moving-baseline solution. The first step is to unify the time and space systems, then make full use of the multi constellation carrier phase measurements, and finally eliminate the propagation path error and satellite clock error by double difference carrier phase. The moving-baseline length between two antennas is calculated in real time by Kalman filter algorithm, and then compared with the actual train length to realize the train integrity monitoring. © 2021, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Satellites

来源：评论

学校读者我要写书评

暂无评论

A fast and effective maximum power point tracking method under various weather conditions for grid-connected photovoltaic systems

A fast and effective maximum power point tracking method und...

引用

IEEE Conference on Power Electronics and Renewable Energy (CPERE)

作者： Karam Khairullah Mohammed Saad Mekhilef Muhyaddin Rawa Control and Mechatronics Engineering Department School of Electrical Engineering Universiti Teknologi Malaysia Johor Bahru Malaysia School of Software and Electrical Engineering Faculty of Science Engineering and Technology Swinburne University of Technology Hawthorn VIC Australia Smart Grids Research Group Center of Research Excellence in Renewable Energy and power Systems King Abdulaziz University Jeddah Saudi Arabia Department Of Electrical and computer Engineering Faculty of engineering K. A Care Energy Research and Innovation Center King Abdulaziz University Jeddah Saudi Arabia

Photovoltaic systems (PV) are exposed to various weather conditions during the day and getting the maximum power of a PV system is one of its major challenges. While the conventional methods take time to track the maximum power accompanying the fluctuation around the maximum power. Consequently, several metaheuristic algorithms have been presented to overcome the fluctuating around the maximum power. On the other hand optimization methods will search into entire the PV curve to extract the maximum power. This leads to unnecessary search points during the uniform shading conditions which decrease the convergence speed towards the maximum power. In overcoming the tracking time issues as well as the capacity to distinguish between uniform and partial shading conditions, a quick and efficient tracking technique connected to the grid system was presented in this paper. The proposed system, has been tested using MATLAB/SIMULINK, which includes a boost converter to track the maximum power with a sampling time of 0.05 seconds, While a voltage source converter (VSC) is used to transfer the power generated by solar panels to the grid. The simulation results demonstrated that the proposed method was successfully implemented, with an average tracking time of 0.412 s for different weather conditions. Moreover, the comparison with the conventional method of perturb and observe (P& O) has been presented to evaluate the proposed method’s efficiency.

关键词： Maximum power point trackers Renewable energy sources Fluctuations Simulation Optimization methods Power electronics Solar panels

来源：评论

学校读者我要写书评

暂无评论

G-Fuzz: A Directed Fuzzing Framework for gVisor

arXiv

引用

arXiv 2024年

作者： Li, Yuwei Chen, Yuan Ji, Shouling Zhang, Xuhong Yan, Guanglu Liu, Alex X. Wu, Chunming Pan, Zulie Lin, Peng The College of Electronic Engineering National University of Defense Technology Anhui Hefei230000 China The College of Computer Science and Technology Zhejiang University Zhejiang Hangzhou310027 China The School of Software Technology Zhejiang University Zhejiang Hangzhou310027 China Ant Group Zhejiang Hangzhou310063 China College of Computer Science and Technology Zhejiang University Zhejiang Hangzhou310027 China Chinese Aeronautical Establishment and AVIC Artificial Intelligence Research Institute Beijing100029 China

gVisor is a Google-published application-level kernel for containers. As gVisor is lightweight and has sound isolation, it has been widely used in many IT enterprises [1]–[3]. When a new vulnerability of the upstream gVisor is found, it is important for the downstream developers to test the corresponding code to maintain the security. To achieve this aim, directed fuzzing is promising. Nevertheless, there are many challenges in applying existing directed fuzzing methods for gVisor. The core reason is that existing directed fuzzers are mainly for general C/C++ applications, while gVisor is an OS kernel written in the Go language. To address the above challenges, we propose G-Fuzz, a directed fuzzing framework for gVisor. There are three core methods in G-Fuzz, including lightweight and fine-grained distance calculation, target related syscall inference and utilization, and exploration and exploitation dynamic switch. Note that the methods of G-Fuzz are general and can be transferred to other OS kernels. We conduct extensive experiments to evaluate the performance of G-Fuzz. Compared to Syzkaller, the state-of-the-art kernel fuzzer, G-Fuzz outperforms it significantly. Furthermore, we have rigorously evaluated the importance for each core method of G-Fuzz. G-Fuzz has been deployed in industry and has detected multiple serious vulnerabilities. Copyright © 2024, The Authors. All rights reserved.

关键词： C++ (programming language)

来源：评论

学校读者我要写书评

暂无评论

Future Backbone Optical Networks: Fiber Densification Versus Network Densification

Future Backbone Optical Networks: Fiber Densification Versus...

引用

International Symposium on Advanced Networks and Telecommunication Systems (ANTS)

作者： Anuj Agrawal Vimal Bhatia Department of Computer Science and Engineering Indian Institute of Technology Gandhinagar India Department of Electrical Engineering Signals and Software Group Indian Institute of Technology Indore India

ISBN: (数字)9781665448932

ISBN: (纸本)9781665448949

The increasing amount of traffic being aggregated to the terrestrial backbone optical networks (BONs) from the metro/access communication networks has caused capacity crunch in the BONs. Increasing coverage with rural and remote connectivity, wireless network densification, data hungry 5G applications, and increasing datacenters (DCs) are among the major factors that necessitated the deployment of huge amount of optical fibers around the world to connect cell sites, to interconnect DCs, to provide fiber to the home (FTTH) services, among others. Due to massive fiberization in the metro/access networks and hence to handle the increased traffic being aggregated to BONs through them, efforts are being made to increase the capacity of BONs via novel devices, fiber designs, and spectral/spatial flexibility and expansion along with the additional deployment of fibers (referred to as fiber densification) in BONs. In this paper, we focus on the topological aspect of BON and present BON densification as an approach to enable the realization of several advanced optical network technologies in BONs. We analyze the advantages of BON densification considering multicore fibers (MCFs) for which low transmission reach has been the main hindrance towards its deployment in BONs. Simulations indicate that through BON densification, MCF networks can achieve significantly low blocking, and high resiliency against multi-link failures. Furthermore, we highlight several other aspects of future communication networks which indicate that BON densification is a promising approach as compared to the conventional fiber densification.

关键词： Optical fibers Wireless networks Optical fiber networks Optical fiber communication Optical fiber devices Telecommunications Optical fiber subscriber loops

来源：评论

学校读者我要写书评

暂无评论

3D Generative Models: A Survey

arXiv

引用

arXiv 2022年

作者： Shi, Zifan Peng, Sida Xu, Yinghao Geiger, Andreas Liao, Yiyi Shen, Yujun Department of Computer Science and Engineering the Hong Kong University of Science and Technology Hong Kong School of Software Technology Zhejiang University China Department of Information Engineering the Chinese University of Hong Kong Hong Kong Autonomous Vision Group University of Tubingen Tubingen AI Center Germany College of Information Science & Electronic Engineering Zhejiang University China Ant Group China

Generative models aim to learn the distribution of observed data by generating new instances. With the advent of neural networks, deep generative models, including variational autoencoders (VAEs), generative adversarial networks (GANs), and diffusion models (DMs), have progressed remarkably in synthesizing 2D images. Recently, researchers started to shift focus from 2D to 3D space, considering that 3D data is more closely aligned with our physical world and holds immense practical potential. However, unlike 2D images, which possess an inherent and efficient representation (i.e., a pixel grid), representing 3D data poses significantly greater challenges. Ideally, a robust 3D representation should be capable of accurately modeling complex shapes and appearances while being highly efficient in handling high-resolution data with high processing speeds and low memory requirements. Regrettably, existing 3D representations, such as point clouds, meshes, and neural fields, often fail to satisfy all of these requirements simultaneously. In this survey, we thoroughly review the ongoing developments of 3D generative models, including methods that employ 2D and 3D supervision. Our analysis centers on generative models, with a particular focus on the representations utilized in this context. We believe our survey will help the community to track the field’s evolution and to spark innovative ideas to propel progress towards solving this challenging task. Copyright © 2022, The Authors. All rights reserved.

关键词： Deep learning

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：