检索结果-内蒙古大学图书馆

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Ziyu Zhu Wenlei Liu Zhidong Deng Department of Computer Science and Technology Tsinghua University Beijing China State Key Laboratory of Intelligent Technology and Systems Beijing National Research Center for Information Science and Technology (BNRist) Institute for Artificial Intelligence at Tsinghua University (THUAI)

Anomaly detection could be applied in a wide range of fields from industrial scene to medical imaging analysis. Although invertible flow models are developed to accomplish unsupervised anomaly detection, they are usually hard to train and have limited capabilities of accurately modeling the distribution of normal samples. To address this problem, we propose a novel enhanced flow model conditioned on graph representation memory (FlowGRM) for visual surface defect detection. In our FlowGRM, a graph neural network is trained to query graph embedding features from memory bank and incorporate them into a learnable flow model as conditional information. Such memorized conditional information, together with the invertible flow model, is jointly optimized to give rise to a probability density score. Experimental results obtained on the industrial visual anomaly detection datasets including MVTec and MTD show that the proposed model has state-of-the-art performance, which could even reach up to 99.6% and 99.2% AUROC scores on the above two bench-marks, respectively. Meanwhile, it requires fewer parameters and saves over 80% training time compared to existing flow models.

关键词： Training Visualization Time series analysis Signal processing Graph neural networks Acoustics Speech processing

来源：评论

学校读者我要写书评

暂无评论

A Machine-Learning Approach for Children’s Pain Assessments Using Prosodic and Spectral Acoustic Features

A Machine-Learning Approach for Children’s Pain Assessments...

引用

Electrical, computer, Communications and Mechatronics Engineering (ICECCME), International Conference on

作者： Abdulrahman Jamal Sadam Al-Azani Information & Computer Science Department SDAIA-KFUPM Joint Research Center for AI King Fahd University of Petroleum & Minerals Dhahran Saudi Arabia SDAIA-KFUPM Joint Research Center for AI King Fahd University of Petroleum & Minerals Dhahran Saudi Arabia

This paper presents a comprehensive study on the application of Artificial Intelligence (AI) techniques for classifying the reasons behind an infant’s cry. The aim of the study is to develop an effective model that can accurately categorize infants’ cries using Machine Learning (ML) and Deep Learning (DL) classification methods. The research focuses on two distinct datasets, namely DonateACry and BabyCry, in order to explore the best approach for cry classification. The study includes a preprocessing step with data augmentation techniques to address dataset imbalances. Two feature extraction methods are employed: audio spectrograms and a hybrid approach combining prosodic and spectral acoustic features. Different classification models, including K-Nearest Neighbors (K-NN), Dense Artificial Neural Networks (ANNs), and Convolutional Neural Networks (CNNs), are evaluated using the f1 score and accuracy. The results demonstrate that the hybrid approach of prosodic and spectral acoustic features outperforms audio spectrograms, pro-viding a more comprehensive representation of infants’ cries and improving classification accuracy. The research showcases the potential of AI-driven advancements in revolutionizing childcare and parenting practices, contributing to improved well-being and care for infants. Overall, this study demonstrates the effectiveness of AI techniques in understanding infant communication and emphasizes their potential to improve care for infants.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Modernizing Home Protection: An IoT-Driven Approach with Smart Lock and Android Application

Modernizing Home Protection: An IoT-Driven Approach with Sma...

引用

IEEE International Conference on Control System, Computing and Engineering (ICCSCE)

作者： Summaiya Rajput Noureen Talpur Rozan Boudville Ghulam Mustafa E Abro Beenish Talpur Faizan Zahid Department of Telecommunications Engineering Mehran University of Engineering and Technology Jamshoro Pakistan Computer Information Science Department Centre for Research in Data Science (CeRDaS) Universiti Teknologi PETRONAS Seri Iskandar Malaysia Electrical Engineering Studies College of Electrical Engineering Universiti Teknologi MARA Permatang Pauh Pulau Pinang Malaysia Interdisciplinary Research Center for Aviation and Space Exploration King Fahd University of Petroleum and Minerals Dhahran Dammam Kingdom of Saudi Arabia Department of Electrical Electronics and Computer Engineering University of Catania Italy

ISBN: (数字)9798350364507

ISBN: (纸本)9798350364514

In the field of access control locks, conventional methods often require users to download applications with Bluetooth connectivity, and biometric measures such as fingerprints or face recognition that introduce challenges like access to only authorized parties, limited device accessibility and intricate registration processes. To address these challenges, the research paper presents an innovative IoT-based Smart Door Lock System, aiming to redefine access control. Diverging from traditional approaches, this system eliminates the need for application downloads and simplifies the complexities tied to biometrics. The core innovation lies in the simplicity of a user-friendly 4-digit PIN for primary access, prioritizing ease of use. To overcome real-time access constraints, the system introduces One-time Password (OTP) and One Touch Access (OTA) modes. These modes eliminate the intricacies linked with traditional methods, ensuring immediate and hassle-free access for both authorized and unauthorized users.

关键词： Technological innovation Databases Face recognition Process control Passwords Fingerprint recognition Extraterrestrial measurements Real-time systems Pins Reliability

来源：评论

学校读者我要写书评

暂无评论

Pansharpening Methods Based on a Redundant Contourlet Transform

Pansharpening Methods Based on a Redundant Contourlet Transf...

引用

2023 International Conference on Earth Observation and Geo-Spatial Information, ICEOGI 2023

作者： Ourabia, Soumeya Baaziz, Nadia Smara, Youcef University Center of Tipaza -Morsli Abdelah-Institut of Sciences Laboratory of Image Processing and Radiation -USTHB Departement of Electronics Tipaza Algeria University of Quebec in Outaouais - Uqo Research Laboratory in Multimedia Information Department of Computer Science and Engineering GatineauQC Canada University of Sciences and Technology Houari Boumediene -USTHB Laboratory of Image Processing and Radiation Faculty of Electrical Engineering Algiers Algeria

ISBN: (纸本)9798350396720

The pansharpening of high-resolution panchromatic (Pan) image and low-resolution multispectral (MS) images represents an important task in the remote sensing field. It allows the joint exploitation of the information derived from both types of images. However, this may cause a distortion of spectral information or a lack in details and structures. The multiresolution analysis using wavelet transforms has proved its efficiency in the pansharpening domain, due to the representation of the image content over different resolution levels. In image fusion applications, we always need more detail information to be incorporated in the pansharpening procedure in order to produce enhanced results. However, the limitation of the wavelet transform to three types of oriented detail coefficients prevents this need from being met. To overcome this limitation, we propose to use the redundant contourlet transform (RCT) which extracts a richer multiscale directional information from the image. For this purpose, two RCT-based pansharpening methods are introduced and suitable data fusion procedures are described. We conducted several experiments on two different datasets acquired respectively by ALSAT-2A and IKONOS satellites. The visual results as well as quantitative results from evaluation metrics demonstrate the performance of the proposed methods. © 2023 IEEE.

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

Proportional-Integral State Estimation for A Class of Nonlinear Complex Networks with Uncertain Couplings: Tackling Adversarial Replay Attacks

引用

IEEE Transactions on Network science and Engineering 2025年

作者： Fu, Haijing Wang, Zidong Zhao, Di Shen, Bo Donghua University College of Information Science and Technology Shanghai201620 China Ministry of Education Engineering Research Center of Digitalized Textile and Fashion Technology Shanghai201620 China Brunel University London Department of Computer Science Middlesex UxbridgeUB8 3PH United Kingdom University of Shanghai for Science and Technology Business School Shanghai200093 China

This paper addresses the proportional-integral state estimation problem for a class of complex networks with uncertain inner couplings under replay attacks. The inner coupling strengths are subject to fluctuations within a specified interval. A replay attack strategy is implemented by the adversary on the communication channel between the network nodes and the state observers, with the intention of replacing the current measurement data with previously recorded measurement data. The dynamic behavior of these replay attacks is characterized by two factors: one dependent on a stochastic variable and the other on a time-varying parameter. A proportional-integral observer (PIO) is developed to handle the challenges posed by uncertain internal couplings and replay attacks. By employing stochastic analysis and matrix inequality techniques, sufficient conditions are obtained to ensure the exponential mean-square stability and H∞ performance of the estimation error dynamics. Moreover, the PIO gain parameters are determined through the solution of some matrix inequalities. Finally, numerical simulation results are presented to verify the effectiveness of the developed PIO design algorithm. © 2013 IEEE.

关键词： Integral equations

来源：评论

学校读者我要写书评

暂无评论

Enhancing Short-Term Load Forecasting Accuracy in the Power Systems Using a Hybrid CNN-GRU Model

Enhancing Short-Term Load Forecasting Accuracy in the Power ...

引用

International Conference on Electrical, Communication, and computer Engineering (ICECCE)

作者： Muhammad Ahsan Muhammad Osama Aziz Syed Ahmad Saleem Bokhari Muhammad Haris Mansoor Iqbal Kaleem Ullah Department of Computer Science University of Alabama at Birmingham Birmingham AL USA Buildings Energy Research Center University of Engineering and Technology Peshawar Pakistan Harvard Extension School Harvard University Cambridge MA USA Institute for Insight - J Mack Robinson College of Business Georgia State University Georgia USA Department of Computer Science University of Wah Wah Cantt Pakistan Department of Electrical Energy System CASE UET Peshawar Peshawar Pakistan

ISBN: (数字)9798331529437

ISBN: (纸本)9798331529444

In modern power systems, Short-Term Load Forecasting (STLF) is critical in ensuring the dynamic allocation of operational reserves and accurate prediction of electrical load demand. STLF is pertinent to grid operators to ensure system stability, allocate generation resources efficiently, and provide adequate reserves to deal with unexpected fluctuations. Considering the pivotal role of STLF, this paper introduces a hybrid deep learning model for short-term load forecasting, combining Convolutional Neural Networks (CNN) and Gated Recurrent Units (GRU). The model features two parallel 32-filter convolutional layers (3×3 kernel, ReLU) for spatial feature extraction, followed by two series-connected GRU layers (42 units) to capture temporal dependencies. A final dense layer with ReLU activation generates the load forecast. This architecture effectively leverages CNNs for local pattern recognition and GRUs for modeling temporal sequences, offering a robust solution for accurate short-term load forecasting. Therefore, the efficacy of the CNN-GRU model is tested and validated using an extensive dataset sourced from the National Transmission & Dispatch Company (NTDC), Pakistan. The performance analysis reveals that hybrid CNN-GRU models achieve a substantial reduction in Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and Root Mean Square Error (RMSE), with values of 400.33, 2.89, 580.95, for single-step forecasts, respectively, and 1031.11, 5.31, and 733.19, for 24-hour forecasts. Furthermore, the simulation results demonstrate notable enhancements in load forecasting accuracy and robustness across various forecast horizons.

关键词： Accuracy Load forecasting Predictive models Power system stability Feature extraction Stability analysis Convolutional neural networks Kernel Forecasting Load modeling

来源：评论

学校读者我要写书评

暂无评论

Parallel Approximations for High-Dimensional Multivariate Normal Probability Computation in Confidence Region Detection Applications

Parallel Approximations for High-Dimensional Multivariate No...

引用

International Symposium on Parallel and Distributed Processing (IPDPS)

作者： Xiran Zhang Sameh Abdulah Jian Cao Hatem Ltaief Ying Sun Marc G. Genton David E. Keyes Computer Electrical and Mathematical Sciences and Engineering Division King Abdullah University of Science and Technology Statistics Program King Abdullah University of Science and Technology Extreme Computing Research Center King Abdullah University of Science and Technology Thuwal Saudi Arabia Department of Mathematics The University of Houston Houston Texas USA

ISBN: (数字)9798350387117

ISBN: (纸本)9798350387124

Addressing the statistical challenge of computing the multivariate normal (MVN) probability in high dimensions holds significant potential for enhancing various applications. For example, the critical task of detecting confidence regions where a process probability surpasses a specific threshold is essential in diverse applications, such as pinpointing tumor locations in magnetic resonance imaging (MRI) scan images, determining hydraulic parameters in groundwater flow issues, and forecasting regional wind power to optimize wind turbine placement, among numerous others. One common way to compute high-dimensional MVN probabilities is the Separation-of-Variables (SOV) algorithm. This algorithm is known for its high computational complexity of O(n 3 ) and space complexity of O(n 2 ), mainly due to a Cholesky factorization operation for an n×n covariance matrix, where n represents the dimensionality of the MVN problem. This work proposes a high-performance computing framework that allows scaling the SOV algorithm and, subsequently, the confidence region detection algorithm. The framework leverages parallel linear algebra algorithms with a task-based programming model to achieve performance scalability in computing process probabilities, especially on large-scale systems. In addition, we enhance our implementation by incorporating Tile Low-Rank (TLR) approximation techniques to reduce algorithmic complexity without compromising the necessary accuracy. To evaluate the performance and accuracy of our framework, we conduct assessments using simulated data and a wind speed dataset. Our proposed implementation effectively handles high-dimensional multivariate normal (MVN) probability computations on shared and distributed-memory systems using finite precision arithmetics and TLR approximation computation. Performance results show a significant speedup of up to 20X in solving the MVN problem using TLR approximation compared to the reference dense solution without sacrificin

关键词： Accuracy Tensors Wind speed Magnetic resonance imaging Linear algebra Wind power generation Approximation algorithms

来源：评论

学校读者我要写书评

暂无评论

Transcriptome Analysis of Bone Marrow Plasma Cells in Multiple Myeloma Patients before Treatment

引用

Biochemistry (Moscow), Supplement Series B: Biomedical Chemistry 2025年第1期19卷 98-108页

作者： Sergeeva, A. M. Gribkova, A. K. Surimova, V. A. Suntsova, M. V. Buzdin, A. A. Shaytan, A. K. National Medical Research Center for Hematology Ministry of Health of Russia Moscow Russia Department of Biology Moscow State University Moscow Russia Digital Biodesign and Personalized Healthcare World-Class Research Center Sechenov First Moscow State Medical University Moscow Russia Moscow Institute of Physics and Technology Dolgoprudny Russia Shemyakin–Ovchinnikov Institute of Bioorganic Chemistry Moscow Russia Laboratory of Bioinformatics Department of Computer Science Higher School of Economics University Moscow Russia

Multiple myeloma (MM) is a malignant lymphoproliferative disorder associated with accumulation of terminally differentiated B lymphocytes (plasma cells) in the bone marrow, monoclonal expression of pathologic immunoglobulin, anemia, renal damage, hypercalcemia, and bone lesions. Despite considerable attention to the study of ММ pathogenesis and the development of new drugs, this disease remains incurable. Omics technologies are contributing significantly to the understanding of the molecular mechanisms of plasma cell neoplastic transformation in MM and may lead to the identification of novel therapeutic targets. In this work, the authors performed comparative gene expression analysis in CD138+ cell samples obtained from bone marrow aspirates of 46 MM patients and seven healthy donors using high-throughput RNA sequencing technology. Differential expression analysis identified 1230 genes with statistically significant expression changes in MM patient samples compared to donor samples. Functional analysis of the transcriptome revealed that pathogenetic changes in MM were associated with groups related to growth factors and intracellular signaling (DKK1, BMP4, HGF, TGFB2, FGF), extracellular matrix modification and regulation of cell adhesion (VCAM1, MMP16, LAMP5), ion channel activity (GRIA3, CLCNKA, GABRB2), regulation of immune functions, chromatin organization, cytoskeleton, and Ca2+ signaling. A significant proportion of genes from the ion channel category were associated with the regulation of neuronal transmission. The last category is poorly characterized, which could provide a new direction for MM therapy. The presented functional analysis of differentially expressed genes helps to elucidate the molecular mechanisms of MM, which will contribute to the development of new treatment approaches.

关键词： transcriptome multiple myeloma plasma cells CD138+cells bone marrow RNA sequencing differential gene expression gene set enrichment analysis

来源：评论

学校读者我要写书评

暂无评论

In situ control and nanofocusing of extreme ultraviolet solid-state high harmonics

In situ control and nanofocusing of extreme ultraviolet soli...

引用

International Conference on Ultrafast Phenomena, UP 2022

作者： Korobenko, Aleksey Rashid, Sabaa Heide, Christian Naumov, Andrei Yu. Reis, David A. Berini, Pierre Corkum, Paul B. Vampa, Giulio Joint Attosecond Science Laboratory National Research Council of Canada University of Ottawa Ottawa Canada Center for Research in Photonics University of Ottawa Ottawa Canada Stanford PULSE Institute SLAC National Accelerator Laboratory Menlo ParkCA94025 United States School of Electrical Engineering and Computer Science University of Ottawa Ottawa Canada Department of Physics University of Ottawa Ottawa Canada

In this work we use helium ion beam nanolithography to structure the surface of an MgO crystal and control the emission of extreme-ultraviolet high harmonics. We demonstrate their focusing to below 300 nm in waist dia... 详细信息

ISBN: (纸本)9781957171074

关键词： Crystal structure

来源：评论

学校读者我要写书评

暂无评论

Fast-Convergent and Communication-Alleviated Heterogeneous Hierarchical Federated Learning in Autonomous Driving

引用

IEEE Transactions on Intelligent Transportation Systems 2025年

作者： Kou, Wei-Bin Lin, Qingfeng Tang, Ming Ye, Rongguang Wang, Shuai Zhu, Guangxu Wu, Yik-Chung The University of Hong Kong Department of Electrical and Electronic Engineering Hong Kong Hong Kong Southern University of Science and Technology Department of Computer Science and Engineering Shenzhen518055 China Chinese Academy of Sciences Shenzhen Institute of Advanced Technology Shenzhen518055 China Shenzhen International Center for Industrial and Applied Mathematics Shenzhen Research Institute of Big Data Shenzhen518172 China

Street Scene Semantic Understanding (denoted as TriSU) is a complex task for autonomous driving (AD). However, inference model trained from data in a particular geographical region faces poor generalization when applied in other regions due to inter-city data domain-shift. Hierarchical Federated Learning (HFL) offers a potential solution for improving TriSU model generalization by collaborative privacy-preserving training over distributed datasets from different cities. Unfortunately, it suffers from slow convergence because the data from different cities are with disparate statistical properties. Going beyond existing HFL methods, we propose a Gaussian heterogeneous HFL algorithm (FedGau) to address inter-city data heterogeneity so that convergence can be accelerated. In the proposed FedGau algorithm, both single RGB image and RGB dataset are modelled as Gaussian distributions for aggregation weight design. This approach not only differentiates each RGB image by respective statistical distribution, but also exploits the statistics of dataset from each city in addition to the conventionally considered data volume. With the proposed approach, the convergence is accelerated by 35.5%-40.6% compared to existing state-of-the-art (SOTA) HFL methods. On the other hand, to reduce the involved communication resource, we further introduce a novel performance-aware adaptive resource scheduling (AdapRS) policy. Unlike the traditional static resource scheduling policy that exchanges a fixed number of models between two adjacent aggregations, AdapRS adjusts the number of model aggregation at different levels of HFL so that unnecessary communications are minimized. Extensive experiments demonstrate that AdapRS saves 29.65% communication overhead compared to conventional static resource scheduling policy while maintaining almost the same performance. © 2000-2011 IEEE.

关键词： Gaussian distribution

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：