检索结果-内蒙古大学图书馆

5th INNS iapr TC 3 GIRPR workshop on artificial neural networks for pattern recognition, ANNPR 2012

作者： Savchenko, Andrey V. National Research University Higher School of Economics Nizhniy Novgorod Russia

ISBN: (纸本)9783642332111

Since the works by Specht, the probabilistic neural networks (PNNs) have attracted researchers due to their ability to increase training speed and their equivalence to the optimal Bayesian decision of classification task. However, it is known that the PNN's conventional implementation is not optimal in statistical recognition of a set of patterns. In this article we present the novel modification of the PNN and prove that it is optimal in this task with general assumptions of the Bayes classifier. The modification is based on a reduction of recognition task to homogeneity testing problem. In the experiment we examine a problem of authorship attribution of Russian texts. Our results support the statement that the proposed network provides better accuracy and is much more resistant to change the smoothing parameter of Gaussian kernel function in comparison with the original PNN. © 2012 Springer-Verlag.

关键词： neural networks

来源：评论

学校读者我要写书评

暂无评论

Capturing Suprasegmental Features of a Voice with RNNs for Improved Speaker Clustering 8th

Capturing Suprasegmental Features of a Voice with RNNs for I...

引用

8th iapr TC3 workshop on artificial neural networks in pattern recognition (ANNPR)

作者： Stadelmann, Thilo Glinski-Haefeli, Sebastian Gerber, Patrick Duerr, Oliver Zurich Univ Appl Sci ZHAW Datalab Winterthur Switzerland Konstanz Univ Appl Sci Inst Opt Syst Constance Germany

ISBN: (纸本)9783319999784;9783319999777

Deep neural networks have become a veritable alternative to classic speaker recognition and clustering methods in recent years. However, while the speech signal clearly is a time series, and despite the body of literature on the benefits of prosodic (suprasegmental) features, identifying voices has usually not been approached with sequence learning methods. Only recently has a recurrent neural network (RNN) been successfully applied to this task, while the use of convolutional neural networks (CNNs) (that are not able to capture arbitrary time dependencies, unlike RNNs) still prevails. In this paper, we show the effectiveness of RNNs for speaker recognition by improving state of the art speaker clustering performance and robustness on the classic TIMIT benchmark. We provide arguments why RNNs are superior by experimentally showing a "sweet spot" of the segment length for successfully capturing prosodic information that has been theoretically predicted in previous work.

关键词： Speaker clustering Speaker recognition Recurrent neural network

来源：评论

学校读者我要写书评

暂无评论

Biologically inspired architecture of feedforward networks for signal classification

Biologically inspired architecture of feedforward networks f...

引用

International Association of pattern recognition, iapr 2000 held jointly with 8th International workshop on Structural and Syntactic pattern recognition, SSPR 2000 and 3rd International workshop on Statistical Techniques in pattern recognition, SPR 2000

作者： Raudys, Šarūnas Tamošiūnaitė, Minija Institute of Mathematics and Informatics Vytautas Magnus University Vileikos 8 KaunasLT-3035 Lithuania

ISBN: (纸本)3540679464

The hypothesis is that in the lowest hidden layers of biological systems "local subnetworks" are smoothing an input signal. The smoothing accuracy may serve as a feature to feed the subsequent layers of the pattern classification network. The present paper suggests a multistage supervised and "unsupervised" training approach for design and training of multilayer feed-forward networks. Following to the methodology used in the statistical pattern recognition systems we split functionally the decision making process into two stages. In an initial stage, we smooth the input signal in a number of different ways and, in the second stage, we use the smoothing accuracy as a new feature to perform a final classification. © Springer-Verlag Berlin Heidelberg 2000.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

Object Detection in Floor Plan Images 8th

Object Detection in Floor Plan Images

引用

8th iapr TC3 workshop on artificial neural networks in pattern recognition (ANNPR)

作者： Ziran, Zahra Marinai, Simone Univ Firenze Dipartimento Ingn Informaz DINFO Florence Italy

ISBN: (纸本)9783319999784;9783319999777

In this work we investigate the use of deep neural networks for object detection in floor plan images. Object detection is important for understanding floor plans and is a preliminary step for their conversion into other representations. In particular, we evaluate the use of object detection architectures, originally designed and trained to recognize objects in images, for recognizing furniture objects as well as doors and windows in floor plans. Even if the problem is somehow easier than the original one in the case of this research the datasets available are extremely small and therefore the training of deep architectures can be problematic. In addition to the use of object detection architectures for floor plan images, another contribution of this paper is the creation of two datasets that have been used for performing the experiments covering different types of floor plans with different peculiarities.

关键词： Floor plan analysis Object detection Convolutional neural networks Transfer learning

来源：评论

学校读者我要写书评

暂无评论

Image pattern recognition in natural environment using morphological feature extraction

Image pattern recognition in natural environment using morph...

引用

作者： Won, Yonggwan Nam, Jiseung Lee, Bae-Ho Department of Computer Engineering Chonnam National University 300 Yongbong-Dong Puk-Gu Kwangju Korea Republic of

ISBN: (纸本)3540679464

The gray-scale morphological Hit-or-Miss transform is theoretically invariant to vertical translation of the input function, which is analogous to gray-value shift of the input images. Designing optimal structuring elements for the Hit-or-Miss transform operator is achieved by neural network learning methodology using a shared-weight neural network (SWNN) architecture. Early stage of the neural network system performs feature extraction using the operator, while the late stage does classification. In experimental studies, this morphological feature-based neural network (MFNN) system is applied to location of human face and automatic recognition of vehicle license plate to examine the property of the operator. The results of the experimental studies show that the gray-scale morphological Hit-or-Miss transform operator is reducing the effects of lighting variation. © Springer-Verlag Berlin Heidelberg 2000.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Emotion recognition in Speech with Deep Learning Architectures 7th

Emotion Recognition in Speech with Deep Learning Architectur...

引用

7th iapr TC3 workshop on artificial neural networks in pattern recognition (ANNPR)

作者： Erdal, Mehmet Kaechele, Markus Schwenker, Friedhelm Univ Ulm Inst Neural Informat Proc D-89081 Ulm Germany

ISBN: (纸本)9783319461823;9783319461816

Deep neural networks (DNNs) became very popular for learning abstract high-level representations from raw data. This lead to improvements in several classification tasks including emotion recognition in speech. Besides the use as feature learner a DNN can also be used as classifier. In any case it is a challenge to determine the number of hidden layers and neurons in each layer for such networks. In this work the architecture of a DNN is determined by a restricted grid-search with the aim to recognize emotion in human speech. Because speech signals are essentially time series the data will be transformed in an appropriate format to use it as input for deep feed forward neural networks without losing much time dependent information. Furthermore the Elman-Net will be examined. The results shows that by maintaining time dependent information in the data better classification accuracies can be achieved with deep architectures.

关键词： Emotion recognition

来源：评论

学校读者我要写书评

暂无评论

Sign-based learning schemes for pattern classification

引用

pattern recognition LETTERS 2005年第12期26卷 1926-1936页

作者： Anastasiadis, AD Magoulas, GD Vrahatis, MN Univ London Birkbeck Coll Sch Comp Sci & Informat Syst Knowledge Lab London WC1N 3QS England Univ London Birkbeck Coll Sch Comp Sci & Informat Syst London WC1E 7HX England Univ Patras UPAIRC Dept Math GR-26110 Patras Greece

This paper introduces a new class of sign-based training algorithms for neural networks that combine the sign-based updates of the Rprop algorithm with the composite nonlinear Jacobi method. The theoretical foundations of the class are described and a heuristic Rprop-based Jacobi algorithm is empirically investigated through simulation experiments in benchmark pattern classification problems. Numerical evidence shows that this new modification of the Rprop algorithm exhibits improved learning speed in all cases tested, and compares favorably against the Rprop and a recently proposed modification, the improved Rprop. (c) 2005 Elsevier B.V. All rights reserved.

关键词： supervised learning nonlinear iterative methods nonlinear Jacobi subminimization feed-forward neural networks Rprop pattern classification

来源：评论

学校读者我要写书评

暂无评论

Improving iris recognition through new target vectors in MLP artificial neural networks

Improving iris recognition through new target vectors in MLP...

引用

5th INNS iapr TC 3 GIRPR workshop on artificial neural networks for pattern recognition, ANNPR 2012

作者： Manzan, José Ricardo Gonçalves Nomura, Shigueo Yamanaka, Keiji Bueno Pereira Carneiro, Milena Veiga, Antônio C. Paschoarelli Faculty of Electrical Engineering Federal University of Uberlândia Campus Santa Mônica Av. João Naves de Ávila 2160 Bloco 3N CEP: 38400-902 - Uberlândia - MG Brazil

ISBN: (纸本)9783642332111

This paper compares the performance of multilayer perceptron (MLP) networks trained with conventional bipolar target vectors (CBVs) and orthogonal bipolar new target vectors (OBVs) for biometric pattern recognition. The experimental analysis consisted of using biometric patterns from CASIA Iris Image Database developed by Chinese Academy of Sciences - Institute of Automation. The experiments were performed in order to obtain the best recognition rates, leading to the comparison of results from both conventional and new target vectors. The experimental results have shown that MLPs trained with OBVs can better recognize the patterns of iris images than MLPs trained with CBVs. © 2012 Springer-Verlag.

关键词： Multilayer neural networks

来源：评论

学校读者我要写书评

暂无评论

A Novel Representation of Graphical patterns for Graph Convolution networks 10th

A Novel Representation of Graphical Patterns for Graph Convo...

引用

10th iapr TC3 workshop on artificial neural networks for pattern recognition (ANNPR)

作者： Benini, Marco Bongini, Pietro Trentin, Edmondo Univ Siena DIISM Siena Italy

ISBN: (纸本)9783031206498;9783031206504

In the context of machine learning on graph data, graph deep learning has captured the attention of many researcher. Due to the promising results of deep learning models in the most diverse fields of application, great efforts have been made to replicate these successes when dealing with graph data. In this work, we propose a novel approach for processing graphs, with the intention of exploiting the already established capabilities of Convolutional neural networks (CNNs) in image processing. To this end we propose a new representation for graphs, called GrapHisto, in the form of unique tensors encapsulating the features of any given graph to then process the new data using the CNN paradigm.

关键词： GrapHisto Graph neural Network Graph Convolution Network

来源：评论

学校读者我要写书评

暂无评论

Parallelized Kernel Patch Clustering

Parallelized Kernel Patch Clustering

引用

4th workshop on artificial neural networks in pattern recognition

作者： Fausser, Stefan Schwenker, Friedhelm Univ Ulm Inst Neural Informat Proc D-89069 Ulm Germany

ISBN: (纸本)9783642121586

Kernel based clustering methods allow to unsupervised partition samples in feature space but have a. quadratic computation time O(n(2)) where n are the number of samples. Therefore these methods are generally ineligible for large datasets. In this paper we propose a meta-algorithm that performs parallelized clusterings of subsets of the samples and merges them repeatedly. The algorithm is able to use many Kernel based clustering methods where we mainly emphasize on Kernel Fuzzy C-Means and Relational neural Gas. We show that the computation time of this algorithm is basicly linear, i.e. O (n). Further we statistically evaluate the performance of this meta-algorithm on a real-life dataset, namely the Enron Emails.

关键词： Cluster analysis

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：