检索结果-内蒙古大学图书馆

IEEE International Conference on Electro-Information technology

作者： David Arnold Jafar Saniie Department of Electrical and Computer Engineering Embedded Computing and Signal Processing (ECASP) Research Laboratory Illinois Institute of Technology Chicago IL U.S.A.

Principal Component Analysis (PCA) is a versatile Unsupervised Learning (UL) technique for reducing the dimensionality of datasets. As a result, PCA is widely used in consumer and research applications as a preprocessing tool for identifying important features prior to further analysis. In instances where on-site personnel or developers do not have the expertise to apply UL techniques, third party processors are frequently retained. However, the release of client or proprietary data poses a substantial security risk. This risk increases the regulatory and contractual burden on analysts when interacting with sensitive or classified information. Homomorphic Encryption (HE) cryptosystems are a novel family of encryption algorithms that permit approximate addition and multiplication on encrypted data. When applied to UL models, such as PCA, experts may apply their expertise while maintaining data privacy. In order to evaluate the potential application of Homomorphic Encryption, we implemented Principal Component Analysis using the Microsoft SEAL HE libraries. The resulting implementation was applied to the MNIST Handwritten dataset for feature reduction and image reconstruction. Based on our results, HE considerably increased the time required to process the dataset. However, the HE algorithm is still viable for non-real-time applications as it had an average pixel error of near-zero for all image reconstructions.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Data Compression for Ultrasonic Microstructure Scattering Signals using Unsupervised Neural Networks

Data Compression for Ultrasonic Microstructure Scattering Si...

引用

IEEE Symposium (IUS) Ultrasonics

作者： Xin Zhang Jafar Saniie Department of Electrical and Computer Engineering Embedded Computing and Signal Processing (ECASP) Research Laboratory Illinois Institute of Technology Chicago Illinois U.S.A.

Ultrasonic Nondestructive Evaluation (NDE) has been extensively used to characterize the microstructure of metallic structures for early exposure of materials integrity. However, industrial NDE requires the processing, storage, and real-time transmission of large volumes of ultrasonic data. Therefore, it is indispensable to compress ultrasonic data with high fidelity. In this study, we explore the development of Unsupervised Learning (UL) based Neural Network (NN) models for massive ultrasonic data compression and an innovative multilayer perceptron residual autoencoder: Ultrasonic Residual Compressive Autoencoder (URCA), is introduced to compress ultrasonic data with high compression performance. This URCA can be fast-trained and utilizes the sparsity penalty with residual connection to optimize compression performance. UL-based NNs allow for memory-efficient training and rapid online augmentation of the model. To examine the results, a high-performance ultrasonic signal acquisition system was assembled to automatically collect ultrasonic data from heat-treated 1,018 steel blocks for microstructure characterization. Compression performance is analyzed based on compression ratio, reconstruction accuracy and model training time. The reconstruction accuracy was measured using the Structural Similarity Index Measure (SSIM) and Peak Signal-to-Noise Ratio (PSNR). By training the URCA NN for a high reconstruction performance of 0.96 SSIM, we obtained 91.25% memory space-saving. For a higher compression performance of 0.80 SSIM, we obtained 96.04% memory space-saving.

关键词：

来源：评论

学校读者我要写书评

暂无评论

IoT-Enabled Smart Bike Helmet with an AI-Driven Collision Avoidance System

IoT-Enabled Smart Bike Helmet with an AI-Driven Collision Av...

引用

IEEE International Conference on Electro-Information technology

作者： Jacob Solus Maureen Rakotondraibe Xinrui Yu Won-Jae Yi Mikhail Gromov Jafar Saniie Department of Electrical and Computer Engineering Embedded Computing and Signal Processing (ECASP) Research Laboratory Illinois Institute of Technology Chicago IL U.S.A.

This paper presents a system design for a smart bike helmet with multiple safety features that are intended to empower bicycle riders to proactively avoid potential sources of danger or injury. A Smart Sensor/Actuator Node (SSAN), driven by an Arduino Uno single-board microcontroller, contains input sensors and actuators to provide riders the ability to send and receive warnings promptly on their helmet. A Vision Node, driven by an NVIDIA Jetson Nano and a cable pin-connected camera, executes AI object detection algorithms for any dangerous objects that are out of sight of the rider and sends alerts to the SSAN as needed. By combining safety features of the SSAN and Vision Node while continuously sending data to an IoT-enabled backend web server, the safety operation of a typical bike ride can be substantially improved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Parallel Read-Write Circuit With Fast Amplitude-Adaptive Matching Scheme to Memristor Crossbar Array

引用

IEEE Transactions on Circuits and Systems I: Regular Papers 2025年

作者： Jiang, Ningye Jiang, Mingxuan Huang, Haoen Yang, Le Huang, Yi Zeng, Zhigang Huazhong University of Science and Technology School of Artificial Intelligence and Automation Key Laboratory of Image Information Processing and Intelligent Control Ministry of Education of China Wuhan430074 China Wuhan Institute of Technology School of Electrical and Information Engineering Wuhan430205 China University of Massachusetts Amherst Department of Electrical and Computer Engineering AmherstMA01003 United States

Memristor crossbar array (MCA) is a computing-in-memory (CIM) module for computational acceleration. However, conventional read-write (R-W) circuits for MCA rely heavily on external components and have shortcoming in long writing times. To address these issues, we propose a parallel R-W circuit with a fast amplitude-adaptive matching (AAM) scheme. The fast AAM scheme is designed to accelerate time of writing memristors in MCA by adaptively adjusting the writing voltage to match an optimal amplitude. The architecture of parallel R-W circuit are outlined with the implementation of R-W processes. The design principles of parallel R-W unit, fast AAM unit, and associated control logic are described for analyzing their functions within the circuit. By integrating the parallel R-W circuit with MCA, a computing block is leveraged in a neural network for classification tasks. The experimental results demonstrate that the neural network with multiple computing blocks maintains high accuracy across various datasets (over 80%). Furthermore, the proposed AAM scheme achieves an average 54% reduction in writing time compared to that one without AAM scheme. Compared to the pulse-width scheme, the parallel R-W circuit exhibits an average 3.05× speedup in adjusting a 5×5 MCA, and the speedup efficiency is higher as the increases of MCA size. © 2025 IEEE.

关键词： Memristors

来源：评论

学校读者我要写书评

暂无评论

On the Impact of Spatial Mode Dispersion for Strongly-Coupled Multicore Fiber Submarine Transmission

On the Impact of Spatial Mode Dispersion for Strongly-Couple...

引用

2023 Optical Fiber Communications Conference and Exhibition, OFC 2023

作者： Sun, Lin Chen, Bin Liu, Gordon Ning Cai, Yi Li, Zhaohui Lu, Chao Shen, Gangxiang Soochow University Suzhou Key Laboratory of Advanced Optical Communication Network Technology School of Electronic and Information Engineering Suzhou215006 China Hefei University of Technology School of Computer Science and Information Engineering Hefei China Sun Yat-Sen University Guangdong Provincial Key Laboratory of Optoelectronic Information Processing Chips School of Electronics and Information Technology Guangzhou511400 China The Hong Kong Polytechnic University Photonics Research Center Department of Electrical and Information Engineering Kowloon Hung Hom Hong Kong

Strongly-coupled multicore fibers exhibit the improved tolerance to fiber non-linearity. Their potentials in optical submarine communications are investigated with considering the coupling length and spatial mode disp... 详细信息

ISBN: (纸本)9781957171180

关键词： Submarines

来源：评论

学校读者我要写书评

暂无评论

Spatial and Contextual Path Network for Image Inpainting

引用

Intelligent Automation & Soft Computing 2024年第2期39卷 115-133页

作者： Dengyong Zhang Yuting Zhao Feng Li Arun Kumar Sangaiah Hunan Provincial Key Laboratory of Intelligent Processing of Big Data on Transportation Changsha University of Science and TechnologyChangsha410114China School of Computer and Communication Engineering Changsha University of Science and TechnologyChangsha410114China International Graduate Institute of AI National Yunlin University of Science and TechnologyYunlinTaiwan Department of Electrical and Computer Engineering Lebanese American UniversityByblosLebanon

Image inpainting is a kind of use known area of information technology to repair the loss or damage to the *** feature extraction is the core of image *** enough space for information and a larger receptive field is very important to realize high-precision image ***,in the process of feature extraction,it is difficult to meet the two requirements of obtaining sufficient spatial information and large receptive fields at the same *** order to obtain more spatial information and a larger receptive field at the same time,we put forward a kind of image restoration based on space path and context path *** the space path,we stack three convolution layers for 1/8 of the figure,the figure retained the rich spatial *** the context path,we use the global average pooling layer,where the accept field is the maximum of the backbone network,and the pooling module can provide global context information for the maximum accept *** order to better integrate the features extracted from the spatial and contextual paths,we study the fusion module of the two *** fusionmodule first path output of the space and context path,and then through themass normalization to balance the scale of the characteristics,finally the characteristics of the pool will be connected into a feature vector and calculate the weight *** of images in order to extract context information,we add attention to the context path refinement *** modules respectively from channel dimension and space dimension to weighted images,in order to obtain more effective *** show that our method is better than the existing technology in the quality and quantity of themethod,and further to expand our network to other inpainting networks,in order to achieve consistent performance improvements.

关键词： Image inpainting attention deep learning convolutional network

来源：评论

学校读者我要写书评

暂无评论

Study of the spectrum effect on the threshold of resonant magnetic perturbation penetration on J-TEXT

引用

plasma Science and technology 2022年第12期24卷 16-25页

作者： Feiyue MAO Nengchao WANG Zhuo HUANG Zhengkang REN Song ZHOU Chengshuo SHEN Xiaoyi ZHANG Ying HE Qi ZHANG Ruo JIA Chuanxu ZHAO Yangbo LI Bo HU Da Li Abba Alhaji BALA Zhipeng CHEN Zhongyong CHEN Zhoujun YANG Yunfeng LIANG Yonghua DING Yuan PAN the J-TEXT Team International Joint Research Laboratory of Magnetic Confinement Fusion and Plasma Physics State Key Laboratory of Advanced Electromagnetic Engineering and TechnologySchool of Electrical and Electronic EngineeringHuazhong University of Science and TechnologyWuhan 430074People's Republic of China College of Computer Science South-Central Minzu UniversityWuhan 430074People's Republic of China Physics Department Federal University DutseDutse 720101JigawaNigeria Forschungszentrum Jülich GmbH Institut für Energie-und Klimaforschung-PlasmaphysikPartner of the Trilateral Euregio Cluster(TEC)Jülich 52425Germany Institute of Plasma Physics Chinese Academy of SciencesHefei 230031People's Republic of China

The spectrum effect on the penetration of resonant magnetic perturbation(RMP) is studied with upgraded in-vessel RMP coils on *** poloidal spectrum of the RMP field,especially the amplitudes of 2/1 and 3/1 components,can be varied by the phase difference between the upper and lower coil rows,ΔΦ=Φ_(top)-Φ_(bottom),where Φ_(top)and Φ_(bottom)are the toroidal phases of the n=1 field of each coil *** type of RMP penetration is found to be related to ΔΦ,including the RMP penetration of either 2/1 or 3/1 RMP and the successive penetrations of 3/1 RMP followed by the 2/1 *** cases with penetration of only one RMP component,the penetration thresholds measured by the corresponding resonant component are close for variousΔΦ.However,the 2/1 RMP penetration threshold is significantly reduced if the 3/1 locked island is formed in *** changes in the rotation profile due to 3/1 locked island formation could partially contribute to the reduction of the 2/1 thresholds.

关键词： tokamak error field resonant magnetic perturbation mode penetration spectrum analysis

来源：评论

学校读者我要写书评

暂无评论

Indoor Navigation System for Visually Impaired People using computer Vision

Indoor Navigation System for Visually Impaired People using ...

引用

2022 IEEE International Conference on Electro Information technology, eIT 2022

作者： Fraga, Adrian Leyte Yu, Xinrui Yi, Won-Jae Saniie, Jafar Illinois Institute of Technology Embedded Computing and Signal Processing Department of Electrical and Computer Engineering Research Laboratory ChicagoIL United States Universidad Politécnica de Madrid E.T.S.I. Industriales Madrid Spain

ISBN: (纸本)9781665480093

This paper studies an indoor navigation guidance system for visually impaired people using Artificial Intelligence (AI) and computer vision techniques to guide users via optimal path based on quick response (QR) code markers and collision avoidance system supported by the monocular depth estimation algorithm. The proposed system utilizes a set of QR code markers, as location beacons, to generate an optimal path for the user to reach the destination. The identified QR code markers are used to correlate and compare information from the online database containing valuable information such as current location information, and the next possible locations from the current location. In addition, our system embraces a safety feature, a collision avoidance system, by utilizing the monocular depth estimation algorithm to identify obstacles in the path to the desired destination. All obtained information is provided to the user through a text-to-speech engine, where the system can direct the user to the optimal path via audio output. © 2022 IEEE.

关键词： Location

来源：评论

学校读者我要写书评

暂无评论

Efficient Fault Detection in Nonlinear Industrial Process: A Reduced Kernel PCA-Based Spectral Clustering Approach

Efficient Fault Detection in Nonlinear Industrial Process: A...

引用

Artificial Intelligence, Applications, Innovation and Ethics (AI2E), International Conference for

作者： Khadija Attouri Majdi Mansouri Mansour Hajji Kais Bouzrara Laboratory of Advanced Materials and Nanotechnologies Higher Institute of Applied Sciences and Technology of Kasserine Kairouan University Tunisia Department of Electrical and Computer Engineering College of Engineering Sultan Qaboos University Muscat Oman Research Laboratory of Automation Signal Processing and Image National Engineering School of Monastir Tunisia

ISBN: (数字)9798331513733

ISBN: (纸本)9798331513740

The challenges posed by nonlinearities in industrial systems necessitate innovative techniques that outperform the limitations of traditional methods such as principal component analysis (PCA). While Kernel Principal Component Analysis (KPCA) offers a robust solution to handle nonlinear data, its computational requirements are a significant issue, especially for large-sized datasets. In this work, we propose a novel technique, namely, reduced kernel principal component analysis-based spectral clustering (RKPCA SpC ), to monitor and detect faults in the benchmark Tennessee Eastman process. The suggested approach addresses the complexity associated with KPCA by reducing data size during the model training phase. This reduction involves retaining only the principal components, preserving informative features, and selecting pertinent samples without compromising the original data's content. The efficacy of the proposed method is evaluated through key performance metrics, including false alarm rate (FAR), missed detection rate (MDR), detection time delay (DTD), and computation time (CT). Additionally, gained execution time (GET), gained storage space (GSP), and loss function (LF) are considered, providing a comprehensive assessment of the developed paradigms' effectiveness. The results demonstrate the promising capabilities of our proposed scheme.

关键词： Training Measurement Technological innovation Fault detection Delay effects Computational modeling Complexity theory Kernel Monitoring Principal component analysis

来源：评论

学校读者我要写书评

暂无评论

Design Flow and Implementation of an AI-Enabled Smart Guidance System for Art Galleries

Design Flow and Implementation of an AI-Enabled Smart Guidan...

引用

2022 IEEE International Conference on Electro Information technology, eIT 2022

作者： Esteban, Alvaro Belloso Delgado, Mario Silvestre Yu, Xinrui Yi, Won-Jae Saniie, Jafar Illinois Institute of Technology Embedded Computing and Signal Processing Department of Electrical and Computer Engineering Research Laboratory ChicagoIL United States Universidad Politécnica de Madrid XE.T.S.I. Industriales Madrid Spain

ISBN: (纸本)9781665480093

Audio guides are commonly utilized to enrich the experience of art gallery visitors and to fully engage them with the artwork by providing background, contexts, and other information related to the corresponding artists. However, this method may be monotonous to the public and cannot automatically change to the appropriate recording when users move to the next artwork. The traditional operation of audio guides is only enabled by the user's visual recognition of the artwork to activate appropriate commentaries manually. In this paper, we explore the potential enhancement of the art gallery visitors' experience by applying Artificial Intelligence (AI), machine learning, and computer vision techniques to recognize the artwork's main features automatically and provide the corresponding artist and other information to the visitors. We train a model using a Deep Convolutional Neural Network (CNN) and Kaggle's Painter by Numbers dataset, and host the trained model on a low-power AI accelerator equipped with a camera to recognize the artwork in real-time. The results are then transmitted to a database, which is utilized by a custom Android application on the smartphone to provide updated artworkrelated information on the display or by audio. Our system targets not only the general public for an interactive experience in the art gallery but also visually impaired people for proper guidance with the smartphone's camera and its voice feedback capabilities. © 2022 IEEE.

关键词： Deep neural 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：