检索结果-内蒙古大学图书馆

international conference on Image and Graphics (ICIG)

作者： Huanhuan Ao Nenghai Yu

Edge is a basic feature in the field of computing vision. So to find edge saliency map is an indispensable operation for many applications on image processing. In this paper we present a fast algorithm to find edge saliency map for a natural image. The approach integrates three basic edge features: edge gradient value, edge segment length and edge density, and it works well to detect salient region boundaries and to suppress ill edges from texture. An edge saliency map can be used to image segmentation and boundaries detection. Experimental results demonstrate that our algorithm outperforms other edge saliency detection methods. Finally, our algorithm is applied on salient objects segmentation, compared with several state of-the-art salient region detection methods and the results show our work is valuable.

关键词： Image edge detection Image segmentation Visualization Feature extraction Computer vision Object segmentation pattern recognition

来源：评论

学校读者我要写书评

暂无评论

Method of Laplacian Eigenmap-Based pattern recognition and Diagnosis for Incipient Fault of pipelines

Method of Laplacian Eigenmap-Based Pattern Recognition and D...

引用

2011 international conference on Internet computing and Information Services （ICICIS 2011）

作者： Zhigang LOU Hongzhao Liu The Faculty of Mechanical and Precision Instrument Engineering Xi'an University of Technology

There is a considerable noise in the measured signal of pressure and flow of a running pipeline due to friction drag and medium diffusion, which poses an obstacle to the quick detection and precise classification of pipeline leakage, especially to the acquiring of weak incipient fault. This paper offers an incipient fault detection method based on nonlinear manifold learning algorithm, which treats the negative pressure wave signal as transient signal and reduces noise of original signal by using multi-scale wavelet transform. The method also learns original fault signal and extracts the intrinsic manifold features of data by using a nonlinear dimensionality reduction algorithm based on Laplacian Eigenmaps. With this method, the identification efficiency of optimal fault characteristics is noticeably improved, and the advantage of this method has been proved by simulation experiments.

关键词： pipeline fault diagnosis Laplacian Eigenmaps Wavelet transform Noise reduction pattern recognition

来源：评论

学校读者我要写书评

暂无评论

Metabolic pathway inference from time series data: A non iterative approach

Metabolic pathway inference from time series data: A non ite...

引用

6th IAPR international conference on pattern recognition in Bioinformatics, PRIB 2011

作者： Astola, Laura Groenenboom, Marian Gomez Roldan, Victoria Van Eeuwijk, Fred Hall, Robert D. Bovy, Arnaud Molenaar, Jaap Biometris Wageningen University and Research Centre Droevendaalsesteeg 1 Wageningen Netherlands Netherlands Consortium for Systems Biology Amsterdam Netherlands Plant Research International Wageningen University Research Centre Wageningen Netherlands Centre for Biosystems Genomics P.O. Box 98 Wageningen Netherlands

ISBN: (纸本)9783642248542

In this article, we present a very fast and easy to implement method for reconstruction of metabolic pathways based on time series data. To model the metabolic reactions, we use the well-established setting of ordinary differential equations. In the present article we consider a network leading to the accumulation of quercetin-glycosides in tomato (Solanum lycopersicum). Quercetin belongs to a group of plant secondary metabolites, generally referred to as flavonoids, which are extensively being studied for their variety of important functions in plants as well as for their potentially health-promoting effects on human. We use time series measurements of metabolite concentrations of quercetin derivatives. In the present setting, the observed concentrations are the variables and the reaction rates are the unknown parameters. A standard method is to solve the parameters by reverse engineering, where the ordinary differential equations (ODE) are solved repeatedly, resulting in impractical computation times. We use an alternative method that estimates the parameters by least squares minimization, and which is, in the order of hundred times faster than the iterative method. Our reconstruction method can incorporate an arbitrary a priori known network structure as well as positivity constraints on the reaction rates. In this way we can avoid over-fitting, which is another often encountered problem in network reconstruction, and thus obtain better estimates for the parameters. We test the presented method by reconstructing artificial networks and compare it with the more conventional method in terms of residuals between the observed and fitted concentrations, computing times and the proportion of correctly identified edges in the network. Finally we exploit this fast method to statistically infer the kinetic constants in the flavonoid pathway. We remark that the method as such is not limited to metabolic network reconstructions, but can be used with any type of time-series

关键词： Ordinary differential equations

来源：评论

学校读者我要写书评

暂无评论

A low-order model of biological neural networks for hierarchical or temporal pattern clustering, detection and recognition

A low-order model of biological neural networks for hierarch...

引用

international Joint conference on Neural Networks (IJCNN)

作者： James Ting-Ho Lo Department of Mathematics and Statistics University of Maryland Baltimore County Baltimore MD USA

A low-order model (LOM) of biological neural networks, which is biologically plausible, is herein reported. LOM is a recurrent hierarchical network composed of novel models of dendritic trees for encoding information, spiking neurons for computing subjective probability distributions and generating spikes, nonspiking neurons for transmitting inhibitory graded signals to modulate their neighboring spiking neurons, unsupervised and supervised covariance learning and accumulation learning mechanisms, synapses, a maximal generalization scheme, and feedback connections with different delay durations. An LOM with a main network that learns without supervision and clusters similar patterns, and offshoot structures that learn with supervision and assign labels to clusters formed in the main network is proposed as a learning machine that learns and retrieves easily, generalizes maximally on corrupted, distorted and occluded temporal and spatial patterns, and utilizes fully the spatially and temporally associated information.

关键词： Neurons Biological system modeling Covariance matrix Biological information theory Machine learning Integrated circuit modeling Biological neural networks

来源：评论

学校读者我要写书评

暂无评论

Optimal distance metric function with trigram features for case based word sense disambiguation using artificial neural network

Optimal distance metric function with trigram features for c...

引用

international conference on Advanced computing, ICAC

作者： P. Tamilselvi S. K. Srivatsa Department of Computer Applications Sathyabama University Chennai India Department of Electronics and Communication Engineering Saint Joseph's College of Engineering Chennai India

ISBN: (纸本)9781467306706

In general, different levels of knowledge are used for disambiguation. In this paper, only three knowledge features or sources (trigram) are used to achieve the word sense disambiguation. Case based approach is applied for the disambiguation process. Cases are nothing but the refined form of words collected from Semcor, used for deriving the sense of the ambiguous input word. All possible Part of Speech (PoS) listed in Brown Corpus are collected and grouped into seventeen groups, and each group is assigned with a constant value. Trigram features of input (ambiguous words) as well as cases are represented as vector of size 1×3. Vector values for the ambiguous word and other two neighboring words are taken out from those assigned weights based on their PoS. In this paper ten different distance metric functions are empirically analyzed for improving the accuracy performance of word disambiguation with minimal knowledge sources. Neural Network is used for extracting correct sense of the ambiguous word from the selected minimal distance cases. In this paper, a long sentence is taken to project the performance of disambiguation process. From the result, it is clear that, post-trigramed Hamming function (F9) produced appreciable disambiguation accuracy 78.57% (recognized eleven ambiguous words out of fourteen).

关键词： Vectors Accuracy Artificial neural networks Measurement Computational linguistics Context pattern recognition

来源：评论

学校读者我要写书评

暂无评论

PADS: A pattern-Driven Stencil Compiler-Based Tool for Reuse of Optimizations on GPGPUs

PADS: A Pattern-Driven Stencil Compiler-Based Tool for Reuse...

引用

international conference on Parallel and Distributed Systems (ICPADS)

作者： Dongni Han Shixiong Xu Li Chen Lei Huang Key Laboratory of Computer System and Architecture Institute of Computing Technology Chinese Academy of Sciences China Computer Science Department Prairie View A&M University USA

Stencil computations are core of wide range of scientific and engineering applications. A lot of efforts have been put into improving efficiency of stencil calculations on different platforms, but unfortunately it is not easy to reuse. In this paper we present a pattern-Driven Stencil compiler-based tool and a simple tuning system to reuse those well optimized methods and codes. We also suggest extensions to OpenMP, depicting high-level data structures in order to facilitate recognition of various stencil computation patterns. The PADS allows programmers to rewrite kernel of stencils or reuse source-to-source translator outputs as optimized stencil template codes with related tuning parameters, In addition, PADS consists of a OpenMP to CUDA translator and code generator using optimized template codes. It also obtains architecture-specific parameters to tune stencils across different GPU platforms. To demonstrate our system flexibility and performance portability, we illustrate four different stencil computations, Laplacian operator with Jacobi iterative method, divergence operator, 3 dimension 25 point stencil and a 2D heat equation using ADI method with periodic boundary conditions. PADS succeeds in generating all these four stencil codes using different optimization strategies and delivers a promising performance improvement.

关键词： pattern matching Kernel Tuning Libraries Optimization Generators

来源：评论

学校读者我要写书评

暂无评论

Fuzzy soft Thresholding Based Hybrid Denoising Model

Communications in Computer and Information Science

引用

Communications in Computer and Information Science 2011年 205 M4D卷 1-10页

作者： Roy, Sudipta Sinha, Nidul Sen, Asoke Kr. Department of IT Assam University Silchar 788011 Assam India Electrical Engineering Department NIT Silchar 788010 Assam India Department of Physics Assam University Silchar 788011 Assam India

ISBN: (纸本)9783642240546

This paper proposes a denoising model hybridized using wavelet and bilateral filters with fuzzy soft thresholding. The parameters of the proposed model are optimized with floating point genetic algorithm (FPGA). The model optimized with one image is used as a general denoising model for other images like Lena, Fetus, Ultrasound, Xray, Baboon, and Zelda. The performance of the proposed model is evaluated in denoising images injected with noises in different degrees;moderate, high and very high, and the results obtained are compared with those obtained with similar hybrid model with wavelet soft thresholding. Results demonstrate that the performance of the proposed model in terms of PSNR and IQI in denoising most of the images is far better than those with similar model with wavelet soft thresholding. It has also been observed that the hybrid model with wavelet soft thresholding fails to denoise images with very high degree of noises while the proposed model can still be capable of denoising. © Springer-Verlag Berlin Heidelberg 2011.

关键词： Image denoising

来源：评论

学校读者我要写书评

暂无评论

Clustering Structured Data with the SPARE Library

Clustering Structured Data with the SPARE Library

引用

2011 4th IEEE international conference on Computer Science and Information Technology(ICCSIT 2011)

作者： Guido Del Vescovo Lorenzo Livi Antonello Rizzi Fabio Massimo Frattale Mascioli Department of Information Engineering Electronics and Telecommunications SAPIENZA University of Rome Via Eudossiana 1800184 RomeItaly

ISBN: (纸本)9781612848334

In this paper we propose the SPARE C++ Library as a flexible tool for solving data driven modelling problems,where the domain space is not necessary constrained to be the set of real valued *** possibility to face pattern recognition problems directly on structured domains (multimedia data,strings,graphs) is fundamental to the effective solution of many interesting applications in the field of content based retrieval and Knowledge *** an instance of this particular characteristic of SPARE,we considered a clustering problem defined in the string domain,focusing on the problem of cluster representation in data domains where only a dissimilarity measure can be *** this aim we propose to adopt the MinSOD (Minimum Sum of Distances) defined as the element of a cluster minimizing the sum of dissimilarities from all the other elements in the considered *** the precise computation of the MinSOD have a high computational cost,we propose a suboptimal procedure consisting in computing the representative of the cluster considering only a reduced pool of samples,instead of the whole set of objects in the *** have carried out some tests in order to ascertain the sensitivity of the clustering procedure with respect to the number of samples in the pool used to compute the *** show a good robustness of the proposed *** implementation is available as part of the SPARE library,published as an open source project.

关键词： pattern recognition SPARE Library Template Meta-programming Clustering Strings k-means MinSOD.

来源：评论

学校读者我要写书评

暂无评论

Warp that smile on your face: Optimal and smooth deformations for face recognition

Warp that smile on your face: Optimal and smooth deformation...

引用

international conference on Automatic Face and Gesture recognition

作者： Tobias Gass Leonid Pishchulin Philippe Dreuw Hermann Ney Human Language Technology and Pattern Recognition Group RWTH Aachen University Germany Computer Vision Laboratory ETH Zurich Switzerland Computer Vision and Multimodal Computing MPI Informatics Saarbruecken Germany

In this work, we present novel warping algorithms for full 2D pixel-grid deformations for face recognition. Due to high variation in face appearance, face recognition is considered a very difficult task, especially if only a single reference image, for example a mug-shot, per face is available. Usually model-based approaches with additional training data are used to cope with several types of variation occurring in facial imaging. Image warping contrarily yields a distance measure which is invariant with regard to several types of variation. This allows for precise recognition even using only very few reference observations. Due to the computationally complex problem of optimal 2D warping, pseudo-2D warping-based approaches in the past represented strong approximations of the original problem, and were mainly successful on data with low variability or rectified images. We propose a novel 2D warping method which is globally optimal and makes no prior assumtions on the data variability besides two-dimensional smootheness constraints which both avoid local mirroring and gaps and significantly speed up the optimization. Furthermore, we show that occlusion handling is imperative to obtain smooth warpings in a variety of domains. We evaluate our novel algorithm on various well known databases, such as the AR-Face and CMU-PIE database, and provide a detailed comparison to existing warping approaches. We show that by using simple relative 2D constraints, strong local features and a kernel, which is robust w.r.t. occlusions, our computationally complex approaches outperform state-of-the-art results for recognizing faces under varying expressions, occlusions and poses. Most interestingly, we achieve higher accuracy using fewer training instances per class compared to methods learning a model of the 3D shape.

关键词： Pixel Face Face recognition Optimization Databases Approximation methods Hidden Markov models

来源：评论

学校读者我要写书评

暂无评论

Small Target Detection Using Motion Compensation and Filtering

Small Target Detection Using Motion Compensation and Filteri...

引用

2011 international conference on Fuzzy Systems and Neural computing(FSNC 2011)

作者： Zhonghua Wang He Deng Jianguo Liu Dengwei Wang School of Information Engineering Nanchang Hangkong University Institute for Pattern Recognition and Artificial Intelligence Huazhong University of Science and Technology

In this paper,we present a novel method which is based on the Entropy flow for detecting moving *** algorithm consists of the following steps:Firstly,2-dimensional histogram of entropy flow field is computed to estimate the background ***,the difference image through background motion compensation is ***,the targets are detected by spatial-temporal *** experiment results demonstrate that the proposed algorithm is robust to noise,and also fit to detect small targets under moving backgrounds in infrared image sequences.

关键词： Entropic map,Entropy flow,Image motion,Moving target

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：