检索结果-内蒙古大学图书馆

Computationally efficient multiscale estimation of large-scale dynamic systems

IEEE International Conference on Image Processing

作者： T.T. Ho P.W. Fieguth A.S. Willsky Lab. for Inf. & Decision Syst. MIT Cambridge MA USA Department of Systems Design Engineering University of Waterloo Canada

ISBN: (纸本)0818688211

Statistical estimation of large-scale dynamic systems governed by stochastic partial differential equations is important in a wide range of scientific applications. However, the realization of computationally efficient algorithms for statistical estimation of such dynamic systems is very difficult. Conventional linear least squares methods are impractical for both computational and storage reasons. A previously-developed multiscale estimation methodology has been successfully applied to a number of large-scale static estimation problems. In this paper we apply the multiscale approach to the more challenging dynamic estimation problems, introducing a recursive procedure that efficiently propagates multiscale models for the estimation errors in a manner analogous to, but more efficient than, the Kalman filter's propagation of the error covariances. We illustrate our research in the context of 1-D and 2-D diffusive processes.

关键词： Large-scale systems Estimation error Sea measurements Stochastic systems Partial differential equations Recursive estimation Kalman filters Optical filters Optical computing Oceans

来源：评论

学校读者我要写书评

暂无评论

COVIDx CXR-3: A Large-Scale, Open-Source Benchmark Dataset of Chest X-ray Images for Computer-Aided COVID-19 Diagnostics

arXiv

引用

arXiv 2022年

作者： Pavlova, Maya Tuinstra, Tia Aboutalebi, Hossein Zhao, Andy Gunraj, Hayden Wong, Alexander Vision and Image Processing Lab University of Waterloo Canada Department of Systems Design Engineering University of Waterloo Canada Cheriton School of Computer Science University of Waterloo Canada Waterloo AI Institute University of Waterloo Canada DarwinAI Corp Canada

After more than two years since the beginning of the COVID-19 pandemic, the pressure of this crisis continues to devastate globally. The use of chest X-ray (CXR) imaging as a complementary screening strategy to RT-PCR testing is not only prevailing but has greatly increased due to its routine clinical use for respiratory complaints. Thus far, many visual perception models have been proposed for COVID-19 screening based on CXR imaging. Nevertheless, the accuracy and the generalization capacity of these models are very much dependent on the diversity and the size of the dataset they were trained on. Motivated by this, we introduce COVIDx CXR-3, a large-scale benchmark dataset of CXR images for supporting COVID-19 computer vision research. COVIDx CXR-3 is composed of 30,386 CXR images from a multinational cohort of 17,026 patients from at least 51 countries, making it, to the best of our knowledge, the most extensive, most diverse COVID-19 CXR dataset in open access form. Here, we provide comprehensive details on the various aspects of the proposed dataset including patient demographics, imaging views, and infection types. The hope is that COVIDx CXR-3 can assist scientists in advancing machine learning research against both the COVID-19 pandemic and related diseases. Copyright © 2022, The Authors. All rights reserved.

关键词： COVID-19

来源：评论

学校读者我要写书评

暂无评论

Experiment and simulation of a coil spring for the performance evaluation of a gas insulated circuit breaker 3

Experiment and simulation of a coil spring for the performan...

引用

3rd International Conference on Electric Power Equipment - Switching Technology, ICEPE-ST 2015

作者： Lee, Dae Woo Kim, Min Soo Kim, Jae Yeol Kim, Hyun Woo Bae, Byung Tae Yoo, Wan Suk Sohn, Jeong Hyun Pukyung National University Department of Mechanical Design Engineering D-BaSE Lab Pusan Korea Republic of Hyosung Corporation Power Systems Performance Unit H.V Switchgear Advanced RandD Team Changwon Korea Republic of Pusan National University School of Mechanical Engineering CAE Labortory Pusan Korea Republic of

ISBN: (纸本)9781467374149

Since the performance of a circuit breaker which has the spring operating mechanism mainly depends on the dynamic behavior of a coil spring, the analysis of the dynamic behavior of a coil spring is required for a performance evaluation of a circuit breaker. When a spring moves so fast from the releasing state, a lot of energy is used to move a self-weight of a spring. In this study, a release test bed is designed for evaluating the dynamic behaviors of a coil spring. From the release tests, the dynamic behaviors of a coil spring are analyzed. RecurDyn is employed to carry out the computer simulation and results are compared with test results. The actual energy transferred to the follower during high-speed motion is analyzed through the comparisons of the kinetic energy between the experiment and the computer simulation. © 2015 IEEE.

关键词： Electric circuit breakers

来源：评论

学校读者我要写书评

暂无评论

Palmprint authentication system for civil applications

Palmprint authentication system for civil applications

引用

8th European Conference on Computer Vision, ECCV 2004 and International Workshop on Biometric Authentication, BioAW 2004

作者： Zhang, David Lu, Guangming Kong, Adams Wai-Kin Wong, Michael Biometrics Research Centre Department of Computing The Hong Kong Polytechnic University Kowloon Hong Kong Biocomputing Research Lab School of Computer Science and Engineering Harbin Institute of Technology Harbin China Department of Systems Design Engineering University of Waterloo OntarioN2L 3G1 Canada

ISBN: (纸本)3540224998

In this paper, we present a novel biometric authentication system to identify a person’s identity by his/her palmprint. In contrast to existing palmprint systems for criminal applications, the proposed system targets at the civil applications, which require identifying a person in a large database with high confidence in real-time. The system is constituted by four major components: User interface Module, Acquisition Module, Recognition Module and External Module. More than 7,000 palmprint images have been collected to test the performance of the system. The system can identify 400 palms with a low false acceptance rate, 0.02%, and a high genuine acceptance rate, 98.83%. For verification, the system can operate at a false acceptance rate, 0.017% and a false rejection rate, 0.86%. The execution time for the whole process including image collection, preprocessing, feature extraction and matching is less than 1 second. © Springer-Verlag Berlin Heidelberg 2004.

关键词： User interfaces

来源：评论

学校读者我要写书评

暂无评论

Saliency-guided compressive fluorescence microscopy

Saliency-guided compressive fluorescence microscopy

引用

Annual International Conference of the IEEE engineering in Medicine and Biology Society

作者： Shimon Schwartz Alexander Wong David A. Clausi Department of Systems Design Engineering University of Waterloo Vision and Image Processing Lab research group Waterloo ON N2L 3G1 Canada

ISBN: (纸本)9781424441198

A novel saliency-guided approach is proposed for improving the acquisition speed of compressive fluorescence microscopy systems. By adaptively optimizing the sampling probability density based on regions of interest instead of the traditional unguided random sampling approach, the proposed saliency-guided compressive fluorescence microscopy approach can achieve high-quality microscopy images using less than half of the number of fluorescence microscopy data measurements required by existing compressive fluorescence microscopy systems to achieve the same level of quality.

关键词： Saliency-guided microscopy high-quality

来源：评论

学校读者我要写书评

暂无评论

Effect of Brightness of Assisted Target Detection Cues in a Simulated Search and Rescue Task

引用

Proceedings of the Human Factors and Ergonomics Society Annual Meeting 2010年第18期54卷 1306 - 1310页

作者： Wayne Giang Allan Keefe Jocelyn Keillor Department of Systems Design Engineering University of Waterloo Waterloo Canada Defence Research and Development Canada - Toronto Toronto Canada Institute for Aerospace Research – Flight Research Lab National Research Council of Canada Ottawa Canada

Assisted target detection (ATD) systems are designed to direct the user's attention to relevant areas of the display, but the majority of the research into the use of such systems does not consider the design of the cue itself. Within a search and rescue (SAR) context, there is a possibility that cues designed to facilitate effective search could in fact distract a SAR operator's search of the terrain, reducing the probability of locating a crashed aircraft. In order to determine if salience matters in the design of an ATD system for video-based sensor systems, it is important to study the impact of highly salient cues on visual search. In a previous experiment where the saliency of a cue was varied using different levels of cue brightness in a search task with static imagery, it was found that the more salient cues produced faster response times without any detrimental effects on accuracy. In the present experiment we used dynamic imagery from a SAR simulator. We found that cues of different brightness improved the sensitivity (d') of participants when compared to conditions in which no cues were available, but there was no evidence of any differences between the different levels of cue brightness. These findings suggest that cue brightness may not influence the salience of cues as much as one might expect in the context of a full-motion simulation. Other visual dimensions such as visual onsets or colour may potentially play a larger role in determining the saliency of ATD system cues when used in a task involving motion such as SAR.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Vision-based speaker detection using Bayesian networks

Vision-based speaker detection using Bayesian networks

引用

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： J.M. Rehg K.P. Murphy P.W. Fieguth Cambridge Research Lab Compaq Computer Corporation Cambridge MA USA Department of Computer Science University of California Berkeley CA USA Department of Systems Design Engineering University of Waterloo Waterloo ONT Canada

The development of user interfaces based on vision and speech requires the solution of a challenging statistical inference problem: The intentions and actions of multiple individuals must be inferred from noisy and ambiguous data. We argue that Bayesian network models are an attractive statistical framework for cue fusion in these applications. Bayes nets combine a natural mechanism for expressing contextual information with efficient algorithms for learning and inference. We illustrate these points through the development of a Bayes net model for detecting when a user is speaking. The model combines four simple vision sensors: face detection, skin color, skin texture, and mouth motion. We present some promising experimental results.

关键词： Bayesian methods Face detection Application software Speech Inference algorithms Skin Computer networks Testing Training data Computer vision

来源：评论

学校读者我要写书评

暂无评论

DenseNet for Breast Tumor Classification in Mammographic Images 1st

DenseNet for Breast Tumor Classification in Mammographic Ima...

引用

1st International Conference on Bioengineering and Biomedical Signal and Image Processing, BIOMESIP 2021

作者： Jiménez Gaona, Yuliana Rodriguez-Alvarez, María José Espino-Morato, Hector Castillo Malla, Darwin Lakshminarayanan, Vasudevan Departamento de Química y Ciencias Exactas Universidad Técnica Particular de Loja San Cayetano Alto s/n CP1101608 Loja Ecuador Valencia46022 Spain Theoretical and Experimental Epistemology Lab School of Optometry and Vision Science Waterloo Canada Department of Systems Design Engineering Physics and Electrical and Computer Engineering University of Waterloo WaterlooONN2L3G1 Canada Valencia46022 Spain

ISBN: (纸本)9783030881627

Breast cancer screening is an efficient method to detect breast lesions early. The common screening techniques are tomosynthesis and mammography images. However, the traditional manual diagnosis requires an intense workload for pathologists, and hence is prone to diagnostic errors. Thus, the aim of this study was to build a deep convolutional neural network method for automatic detection, segmentation, and classification of breast lesions in mammography images. Based on deep learning the Mask-CNN (RoIAlign) method was developed to automate RoI segmentation. Then feature extraction, selection and classification were carried out by the DenseNet architecture. Finally, the precision and accuracy of the model was evaluated by the AUC, accuracy and precision metrics. To summarize, the findings of this study show that the methodology may improve the diagnosis and efficiency in automatic tumor localization through medical image classification. © 2021, Springer Nature Switzerland AG.

关键词： Mammography

来源：评论

学校读者我要写书评

暂无评论

Cross-day analysis of Multicode Surface Electromyography based Biometrics for Personal Identification

Cross-day analysis of Multicode Surface Electromyography bas...

引用

Annual International Conference of the IEEE engineering in Medicine and Biology Society (EMBC)

作者： Ashirbad Pradhan Jiayuan He Ning Jiang Department of Systems Design Engineering Engineering Bionics Lab Faculty of Engineering University of Waterloo Waterloo Canada National Clinical Research Center for Geriatrics West China Hospital Sichuan University China Manufacturing Sichuan University Chengdu China

Recently, surface electromyography (sEMG) has emerged as a novel biometric trait for personal identification, potentially providing a superior spoof-resistant solution over existing traits. The sEMG possesses a unique dual-mode security: they differ between individuals (biometric-mode), and different gestures have different sEMG characteristics (knowledge-mode). To leverage the knowledge-mode facet of the dual-mode security, the previous studies have utilized a multicode framework involving the fusion of codes (gestures), however, the analysis involved data recorded on a single day and from a small subject-pool. In this study, wrist EMG data collected from 43 participants over three different days while performing static hand/wrist gestures was utilized in two cross-day analyses, where the training and testing data were from different days. Three levels of fusion, score, rank, and decision were investigated to determine the optimal fusion scheme. The results showed that the score-level fusion scheme resulted in a median rank-1 accuracy of 77.9% and rank-5 accuracy of 99.6%, all significantly higher (p<0.001) than the respective single-code gesture. Our results showed that the multicode sEMG biometric framework provides superior identification performance in a more realistic cross-day scenario.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Reinforcement learning for parameter control of text detection in images from video sequences

Reinforcement learning for parameter control of text detecti...

引用

International Conference on Information and Communication Technologies: From Theory to Applications, ICTTA

作者： G.W. Taylor C. Wolf PAMI Lab Department of Systems Design Engineering University of Waterloo Waterloo ONT Canada Laboratoire d Informatique en Images et Systèmes dinformation INSA Lyon Villeurbanne France

This work presents a generalized framework for approaching the problem of optimizing parameters for a multistep computer vision algorithm using reinforcement learning (RL) agent. We then focus on the specific task of text detection in images taken from video sequences for semantic indexing. Employing the fuzzy ARTMAP ANN is an effective way to manage and predict large state and action spaces. Image features in the state space will allow for optimization on a per-image.

关键词： Learning Video sequences Computer vision design engineering systems engineering and theory Indexing Extrapolation Image processing Prediction algorithms Multidimensional systems

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：