检索结果-内蒙古大学图书馆

12th International Conference on Communication Networks (ICCN) / 12th International Conference on Data Mining and Warehousing (ICDMW) / 12th International Conference on image and signal processing (ICISP)

作者： Arvind, C. S. Vanjare, Ashoka Omkar, S. N. Senthilnath, J. Mani, V. Diwakar, P. G. Robert Bosch Engn & Business Solut Private Ltd Bangalore Karnataka India Indian Inst Sci Bangalore Karnataka India Indian Space Res Org Bangalore Karnataka India

Flood assessment using unsupervised techniques for multi-temporal MODIS satellite images is presented. Classical methods like mean shift algorithm is compared with artificial neural network method like self organizing maps for automatic water pixel identification and extraction. The extracted results help in identification of flooded and non-flooded places. Different methods are applied and comparative study of unsupervised methods involving mean shift and self organizing maps are carried-out. In order to evaluate the algorithmic performance, root mean square error and receiver operating characteristics is used as performance evaluation indices. The results reported will provide useful information for multi-temporal time series satellite image analysis which can be used for current and future research in disasters management. (C) 2016 The Authors. Published by Elsevier B.V.

关键词： image Segmentation Mean Shift MODIS Satellite images Receiver of Characteristics Root Means Square Self-Organizing Maps Unsupervised Techniques

来源：评论

学校读者我要写书评

暂无评论

Recognition of Faces - An Optimized Algorithmic Chain 12

Recognition of Faces - An Optimized Algorithmic Chain

引用

作者： Shankar, Sheela Udupi, V. R. Dr MS Sheshagiri CET KLE Belgaum 590008 Karnataka India Gogte Inst Technol Belgaum 590008 Karnataka India

Face recognition has been one of the popular and important parts in Human Computer Interaction (HCI) systems that find tremendous applications, some of which are very critical like access control, surveillance, etc. There are numerous techniques available to process face images and hence, choosing an optimal algorithmic chain is not a straight forward job. The scenario gets much more interesting while implementing face recognition in applications connected to cloud via Internet of Things (IoT) platform. This paper reviews some of the effective face recognition algorithms and proposes an optimized algorithmic chain offering optimal classification accuracy and lower execution time;thereby making it appropriate for IoT related applications targeting human-centric systems. Also, achieving optimum efficiency by selecting appropriate number of features for a given combination of algorithms and the behaviour of algorithms due to partitioning of the images in case of Local Binary Pattern (LBP) is discussed. Results indicate enhanced classification rates with algorithmic fusion by creating chains or process flow of methods. Accuracy of up to 96% was obtained for one of the chains that were designed. Also, it is evident from the results that this chain outperforms some of the well-known state-of-the-art methods. (C) 2016 The Authors. Published by Elsevier B.V.

关键词： Face Recognition HCI IoT LBP neural Networks (NN)

来源：评论

学校读者我要写书评

暂无评论

基于无监督学习卷积神经网络的振动信号模态参数识别

引用

计算机应用 2017年第3期37卷 786-790,822页

作者：方宁周宇叶庆卫李玉刚宁波大学信息科学与工程学院浙江宁波315211

针对现有的时域模态参数识别方法大多存在难定阶和抗噪性差的问题,提出一种无监督学习的卷积神经网络(CNN)的振动信号模态识别方法。该算法在卷积神经网络的基础上进行改进。首先,将应用于二维图像处理的卷积神经网络改成处理一维信号... 详细信息

针对现有的时域模态参数识别方法大多存在难定阶和抗噪性差的问题,提出一种无监督学习的卷积神经网络(CNN)的振动信号模态识别方法。该算法在卷积神经网络的基础上进行改进。首先,将应用于二维图像处理的卷积神经网络改成处理一维信号的卷积神经网络,其中输入层改成待提取模态参数的振动信号集合,中间层改成若干一维卷积层、抽样层,输出层得到的为信号对应的N阶模态参数集合;然后,在误差评估中,对网络计算结果(N阶模态参数集)进行振动信号重构;最后,将重构信号和输入信号之间差的平方和作为网络学习误差,使得网络变成无监督学习网络,避免模态参数提取算法的定阶难题。实验结果表明,当所构建的卷积神经网络应用于模态参数提取时,与随机子空间识别(SSI)算法及其局部线性嵌入(LLE)算法对比,在噪声干扰下,构建的卷积神经网络识别精度要高于SSI算法与LLE算法,具有抗噪声强、避免了定阶难题的优点。

关键词：卷积神经网络模态参数无监督学习学习误差随机子空间识别局部线性嵌入

来源：评论

学校读者我要写书评

暂无评论

A new method for extraction of image's features: Complex discrete Ripplet-ii transform

A new method for extraction of image's features: Complex dis...

引用

IEEE signal processing and Communications Applications (SIU)

作者： Hüseyin Yaşar Murat Ceylan T.C Sağlık Bakanlığı Ankara Türkiye KTO Karatay Üniversitesi Mühendislik Fakültesi Elektrik-Elektronik Mühendisliği Bölümü Konya Türkiye

Wavelet transform extracts the features of a signal and image via shifting and weighting methods. This transform has either advantages or disadvantages on image processing applications. One of important disadvantage of wavelet transform is limited orientation problem. This problem has been solved by different orientation with ridgelet transform. Ripplet-ii transform is defined by recently generalising of the ridgelet transform by adding parameter degree (d). Complex discrete form of ripplet-ii transform defined by this study and added to the literature. Also, complex discrete Ripplet-ii transform was tested on medical image classification application. For this, the most common benign lesions in liver MR are classified as cyst and hemangioma using complex discrete Ripplet-ii transform and ANN. Obtained results shown that, classification success of complex discrete Ripplet-ii transform with complex-valued coefficients is higher than real-valued coefficients.

关键词： Radon Feature extraction Wavelet transforms Liver Artificial neural networks

来源：评论

学校读者我要写书评

暂无评论

Deep discriminative manifold learning

Deep discriminative manifold learning

引用

International Conference on Acoustics, Speech, and signal processing (ICASSP)

作者： Jen-Tzung Chien Ching-Huai Chen Department of Electrical and Computer Engineering National Chiao Tung University Hsinchu ROC Taiwan

ISBN: (纸本)9781479999897

This paper presents a new non-linear dimensionality reduction with stochastic neighbor embedding. A deep neural network is developed for discriminative manifold learning where the class information in transformed low-dimensional space is preserved. Importantly, the objective function for deep manifold learning is formed as the Kullback-Leibler divergence between the probability measures of the labeled samples in high-dimensional and low-dimensional spaces. Different from conventional methods, the derived objective does not require the empirically-tuned parameter. This objective is optimized to attractive those samples from the same class to be close together and simultaneously impose those samples from different classes to be far apart. In the experiments on image and audio tasks, we illustrate the effectiveness of the proposed discriminative manifold learning in terms of visualization and classification performance.

关键词： Manifolds Linear programming Neurons Extraterrestrial measurements Principal component analysis Optimization neural networks

来源：评论

学校读者我要写书评

暂无评论

Towards optimal nonlinearities for sparse recovery using higher-order statistics

Towards optimal nonlinearities for sparse recovery using hig...

引用

IEEE Workshop on Machine Learning for signal processing

作者： Steffen Limmer Slawomir Stanczak Network Information Theory Group Technische Universität Berlin Fraunhofer Institute for Telecommunications Heinrich Hertz Institute Berlin Germany

ISBN: (纸本)9781509007479

We consider machine learning techniques to develop low-latency approximate solutions for a class of inverse problems. More precisely, we use a probabilistic approach to the problem of recovering sparse stochastic signals that are members of the l p -balls. In this context, we analyze the Bayesian mean-square-error (MSE) for two types of estimators: (i) a linear estimator and (ii) a structured estimator composed of a linear operator followed by a Cartesian product of univariate nonlinear mappings. By construction, the complexity of the proposed nonlinear estimator is comparable to that of its linear counterpart since the nonlinear mapping can be implemented efficiently in hardware by means of look-up tables (LUTs). The proposed structure lends itself to neural networks and iterative shrinkage/thresholding-type algorithms restricted to a single iteration (e.g. due to imposed hardware or latency constraints). By resorting to an alternating minimization technique, we obtain a sequence of optimized linear operators and nonlinear mappings that converge in the MSE objective. The result is attractive for real-time applications where general iterative and convex optimization methods are infeasible.

关键词： Bayes methods Estimation Probabilistic logic stochastic processes Hardware Standards Higher order statistics

来源：评论

学校读者我要写书评

暂无评论

Biometric Spoofing Detection by a Domain-Aware Convolutional neural Network

Biometric Spoofing Detection by a Domain-Aware Convolutional...

引用

International IEEE Conference on signal-image Technologies and Internet-Based System

作者： Diego Gragnaniello Carlo Sansone Giovanni Poggi Luisa Verdoliva Inst. for High Performance Comput. & Networking Naples Italy DIETI University Federico II Naples Italy

Biometric authentication systems are pervasive in modern society, but they are quite vulnerable to spoofing attacks. Research on spoofing (or liveness) detection is therefore very active. A number of methods have been proposed in the literature, sometimes with very promising results, but limited robustness with respect to the large variety of biometric traits, sensors, and attacks encountered in real-life. Recently, methods based on Convolutional neural Networks (CNNs) are drawing great attention, given their success in many other image processing tasks. However, despite some promising results, they seem to suffer the same robustness problem, requiring heavy training to work properly. Here, we propose a new CNN architecture for biometric spoofing detection. Thanks to domain-specific knowledge, accounted for through a suitable loss function, a compact architecture is obtained, allowing reliable training also in the presence of small-size datasets. Experiments prove the proposal to provide state-of-art performance on several widespread datasets for face and iris liveness detection.

关键词： Feature extraction Convolution Face Training Computer architecture Neurons Biomedical imaging

来源：评论

学校读者我要写书评

暂无评论

Analysis of myocardial texture in resting echocardiographic images predicts recovery one year after myocardial infarction

Analysis of myocardial texture in resting echocardiographic ...

引用

IEEE signal processing Algorithms, Architectures, Arrangements, and Applications Conference

作者： M. Strzelecki S. Skonieczka J. D. Kasprzak M. Simiera B. W. Michalski P. Lipiec Institute of Electronics Technical University of Lodz Lodz Poland II Chair and Dept. of Cardiology Medical University of Lodz Lodz Poland

ISBN: (纸本)9781509026616

The aim of this research is to develop a method for prediction of left ventricular recovery one year after myocardial infarction using texture parameters estimated for static ultrasound images. The study is performed for the monochrome and color (contrast based) echocardiograms that allow advanced evaluation of myocardial function. The analysis includes investigation of different texture feature selection methods and application of multilayer neural network, Support Vector Machine, and decision tree based Adaboost algorithm for recovery prediction. The obtained preliminary results are promising;estimated prediction error is lower than 30% for all used prediction tools.

关键词： Myocardium Support vector machines Decision trees Heart Ultrasonic imaging image color analysis Imaging Ultrasonography Decision Trees Myocardium image colour analysis Support Vector Network Myocardial Infarction Heart Imaging Error of prediction texture Color Contrast

来源：评论

学校读者我要写书评

暂无评论

Fruit classification by HPA-SLFN

Fruit classification by HPA-SLFN

引用

International Conference on Wireless Communications and signal processing (WCSP)

作者： Siyuan Lu Zhihai Lu Preetha Phillips Shuihua Wang Jianguo Wu Yudong Zhang School of Computer Science and Technology Nanjing Normal University Nanjing Jiangsu China School of Natural Sciences and Mathematics Shepherd University Shepherdstown West Virginia USA College of Food Science and Nutritional Engineering China Agricultural University Beijing China

ISBN: (纸本)9781509028610

(Objective) Fruit classification remains a challenge because of the similarities involved by a large quantities of types of fruits. With the aim of recognizing fruits accurately and efficiently, this paper offered a novel fruit-classification tool. (Method) The proposed methodology consisted of following four processes: (i) A four-step preprocessing was performed. (ii) The color, shape, texture features were combined. (iii) Principal component analysis was employed for feature reduction. (iv) We presented a novel classification method with the combination of “Hybridization of PSO and ABC (HPA)” and “single-hidden layer feedforward neural-network (SLFN)”, which was termed as HPA-SLFN. (Results) The experiment results demonstrated that the proposed HPA-SLFN achieved an 89.5% accuracy that was superior to existing methods. (Conclusion) The proposed HPA-SLFN was effective.

关键词： Feature extraction image color analysis Principal component analysis Pipelines Shape Feedforward neural networks Neurons

来源：评论

学校读者我要写书评

暂无评论

stochastic implementation of the disparity energy model for depth perception

Stochastic implementation of the disparity energy model for ...

引用

IEEE Workshop on signal processing Systems (SIPS)

作者： Kaushik Boga Naoya Onizawa François Leduc-Primeau Kazumichi Matsumiya Takahiro Hanyu Warren J. Gross Department of Electrical and Computer Engineering McGill University Montréal Québec Canada Tohuku University Sendai Japan

We implement a binocular vision system based on a disparity-energy model that emulates the hierarchical multi-layered neural structure in the primary visual cortex. Layer 1 performs difference-of-Gaussian filtering that mimicks the center-surround receptive fields (RF) in the retina, layer 2 performs Gabor filtering mimicking the orientation selective filtering performed by simple cells and layer 3 has complex cells tuned to detecting 5 different disparities. A VLSI architecture is developed based on stochastic computing that is compact and adder-free. Even with a short stream length, the proposed architecture achieves better disparity detection than a floating-point version by using a modified disparity-energy model. A 1 × 100 pixel processing system is synthesized using TSMC 65nm CMOS technology and achieves up to 79% reduction in area-delay product compared to a fixed point implementation.

关键词： stochastic processes Approximation methods Brain modeling Computational modeling Computer architecture image coding Streaming media

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：