scholarly journals Constrained K-Means Classification

2018 ◽  
Vol 8 (4) ◽  
pp. 3203-3208
Author(s):  
P. N. Smyrlis ◽  
D. C. Tsouros ◽  
M. G. Tsipouras
Keyword(s):  
Euclidean Distance ◽  
Similarity Criterion ◽  
Experimental Results ◽  
Classification Tasks

Classification-via-clustering (CvC) is a widely used method, using a clustering procedure to perform classification tasks. In this paper, a novel K-Means-based CvC algorithm is presented, analysed and evaluated. Two additional techniques are employed to reduce the effects of the limitations of K-Means. A hypercube of constraints is defined for each centroid and weights are acquired for each attribute of each class, for the use of a weighted Euclidean distance as a similarity criterion in the clustering procedure. Experiments are made with 42 well–known classification datasets. The experimental results demonstrate that the proposed algorithm outperforms CvC with simple K-Means.

Palm Vein Recognition Based on Independent Component Analysis

2013 ◽  
Vol 333-335 ◽  
pp. 1106-1109
Author(s):  
Wei Wu
Keyword(s):  
Pattern Recognition ◽  
Independent Component Analysis ◽  
Euclidean Distance ◽  
Recognition Performance ◽  
Component Analysis ◽  
Independent Component ◽  
Experimental Results ◽  
Vein Pattern ◽  
Vein Recognition ◽  
Palm Vein

Palm vein pattern recognition is one of the newest biometric techniques researched today. This paper proposes project the palm vein image matrix based on independent component analysis directly, then calculates the Euclidean distance of the projection matrix, seeks the nearest distance for classification. The experiment has been done in a self-build palm vein database. Experimental results show that the algorithm of independent component analysis is suitable for palm vein recognition and the recognition performance is practical.

scholarly journals Aspect-augmented Adversarial Networks for Domain Adaptation

2017 ◽  
Vol 5 ◽  
pp. 515-528 ◽  
Author(s):  
Yuan Zhang ◽  
Regina Barzilay ◽  
Tommi Jaakkola
Keyword(s):  
Transfer Learning ◽  
Domain Adaptation ◽  
Experimental Results ◽  
Dependent Manner ◽  
Target Classification ◽  
Adversarial Networks ◽  
Adversarial Training ◽  
Classification Tasks ◽  
Class Labels

We introduce a neural method for transfer learning between two (source and target) classification tasks or aspects over the same domain. Rather than training on target labels, we use a few keywords pertaining to source and target aspects indicating sentence relevance instead of document class labels. Documents are encoded by learning to embed and softly select relevant sentences in an aspect-dependent manner. A shared classifier is trained on the source encoded documents and labels, and applied to target encoded documents. We ensure transfer through aspect-adversarial training so that encoded documents are, as sets, aspect-invariant. Experimental results demonstrate that our approach outperforms different baselines and model variants on two datasets, yielding an improvement of 27% on a pathology dataset and 5% on a review dataset.

scholarly journals Modified Depth-Map Inpainting Method Using the Neural Network

2019 ◽  
Vol 224 ◽  
pp. 04005
Author(s):  
Nikolay Gapon ◽  
Roman Sizyakin ◽  
Marina Zhdanova ◽  
Oksana Balabaeva ◽  
Yigang Cen
Keyword(s):  
Neural Network ◽  
Geometric Model ◽  
Similarity Criterion ◽  
Depth Map ◽  
Experimental Results ◽  
Stereo Pair ◽  
Image Structure ◽  
The Neural Network ◽  
Modern Methods ◽  
Trained Neural Network

This paper proposes a method for reconstructing a depth map obtained using a stereo pair image. The proposed approach is based on a geometric model for the synthesis of patches. The entire image is preliminarily divided into blocks of different size, where large blocks are used to restore homogeneous areas, and small blocks are used to restore details of the image structure. Lost pixels are recovered by copying the pixel values from the source based on the similarity criterion. We used a trained neural network to select the “best like” patch. Experimental results show that the proposed method gives better results than other modern methods, both in subjective and objective measurements for reconstructing a depth map.

Traffic Sign Recognition Based on SIFT Features

2010 ◽  
Vol 121-122 ◽  
pp. 596-599 ◽  
Author(s):  
Ni An Cai ◽  
Wen Zhao Liang ◽  
Shao Qiu Xu ◽  
Fang Zhen Li
Keyword(s):  
Traditional Method ◽  
Euclidean Distance ◽  
Experimental Results ◽  
Traffic Sign Recognition ◽  
Recognition Method ◽  
Traffic Sign ◽  
Traffic Signs ◽  
Sign Recognition ◽  
Good Ability ◽  
Sift Features

A recognition method for traffic signs based on an SIFT features is proposed to solve the problems of distortion and occlusion. SIFT features are first extracted from traffic signs and matched by using the Euclidean distance. Then the recognition is implemented based on the similarity. Experimental results show that the proposed method, superior to traditional method, can excellently recognize traffic signs with the transformation of scale, rotation, and distortion and has a good ability of anti-noise and anti-occlusion.

Reference Point-Based SIFT Feature Matching

2014 ◽  
Vol 543-547 ◽  
pp. 2670-2673
Author(s):  
Lei Cao ◽  
Di Liao ◽  
Bin Dang Xue
Keyword(s):  
Reference Point ◽  
Image Matching ◽  
Euclidean Distance ◽  
Feature Matching ◽  
Experimental Results ◽  
Feature Descriptor ◽  
Matching Algorithm ◽  
Sift Feature ◽  
Sift Features ◽  
Suitable Reference

Aiming to solve the high computational and time consuming problem in SIFT feature matching, this paper presents an improved SIFT feature matching algorithm based on reference point. The algorithm starts from selecting a suitable reference point in the feature descriptor space when SIFT features are extracted. In the feature matching stage, this paper uses the Euclidean distance between descriptor vectors of the feature point to be matched and the reference point to make a fast filtration which removes most of the features that could not be matched. For the remaining SIFT features, Best-bin-first (BBF) algrithm is utilized to obtain precise matches. Experimental results demonstrate that the proposed matching algorithm achieves good effectiveness in image matching, and takes only about 60 percent of the time that the traditional matching algorithm takes.

A Improved Infrared and Visible Images Matching Based on SURF

2013 ◽  
Vol 325-326 ◽  
pp. 1637-1640
Author(s):  
Dong Mei Li ◽  
Jing Lei Zhang
Keyword(s):  
Image Registration ◽  
Euclidean Distance ◽  
Experimental Results ◽  
Interest Points ◽  
Speeded Up Robust Features ◽  
Visible Images ◽  
Data Points

Images matching is the basis of image registration. For their difference, a improved SURF(speeded up robust features) algorithm was proposed for the infrared and visible images matching. Firstly, edges were extracted from the images to improve the similarity of infrared and visible images. Then SURF algorithm was used to detect interest points, and the dimension of the point descriptor was 64. Finally, found the matching points by Euclidean distance. Experimental results show that some invalid data points were eliminated.

Macro Manifold Learning with Applications to Supervised Classification

2014 ◽  
Vol 556-562 ◽  
pp. 3590-3593
Author(s):  
Hong Bing Huang
Keyword(s):  
Pattern Recognition ◽  
Dimensionality Reduction ◽  
Manifold Learning ◽  
Supervised Classification ◽  
Experimental Results ◽  
Classification Tasks ◽  
Low Dimensional

Manifold learning has made many successful applications in the fields of dimensionality reduction and pattern recognition. However, when it is used for supervised classification, the result is still unsatisfactory. To address this challenge, a novel supervised approach, namely macro manifold learning (MML) is proposed. Based on the proposed approach, the low-dimensional embeddings of the testing samples is more favorable for classification tasks. Experimental results demonstrate the feasibility and effectiveness of our proposed approach.

scholarly journals Norm-Explicit Quantization: Improving Vector Quantization for Maximum Inner Product Search

2020 ◽  
Vol 34 (01) ◽  
pp. 51-58 ◽  
Author(s):  
Xinyan Dai ◽  
Xiao Yan ◽  
Kelvin K. W. Ng ◽  
Jiu Liu ◽  
James Cheng
Keyword(s):  
Data Compression ◽  
Vector Quantization ◽  
Similarity Search ◽  
Euclidean Distance ◽  
Quantization Error ◽  
Experimental Results ◽  
Inner Product ◽  
Direction Error ◽  
Product Search ◽  
Norm Error

Vector quantization (VQ) techniques are widely used in similarity search for data compression, computation acceleration and etc. Originally designed for Euclidean distance, existing VQ techniques (e.g., PQ, AQ) explicitly or implicitly minimize the quantization error. In this paper, we present a new angle to analyze the quantization error, which decomposes the quantization error into norm error and direction error. We show that quantization errors in norm have much higher influence on inner products than quantization errors in direction, and small quantization error does not necessarily lead to good performance in maximum inner product search (MIPS). Based on this observation, we propose norm-explicit quantization (NEQ) — a general paradigm that improves existing VQ techniques for MIPS. NEQ quantizes the norms of items in a dataset explicitly to reduce errors in norm, which is crucial for MIPS. For the direction vectors, NEQ can simply reuse an existing VQ technique to quantize them without modification. We conducted extensive experiments on a variety of datasets and parameter configurations. The experimental results show that NEQ improves the performance of various VQ techniques for MIPS, including PQ, OPQ, RQ and AQ.

scholarly journals Selection of Features on Mining Techniques for Classification

2018 ◽  
Vol 7 (S1) ◽  
pp. 108-111
Author(s):  
Gurrampally Kumar ◽  
S. Mohan ◽  
G. Prabakaran
Keyword(s):  
Feature Selection ◽  
Linear Regression ◽  
Gaussian Process ◽  
Statistical Method ◽  
Correlation Coefficient ◽  
Euclidean Distance ◽  
Experimental Results ◽  
Selection Of

Feature selection has been developed by several mining techniques for classification. Some existing approaches couldn’t remove the irrelevant data from dataset for class. Thus it needs the selection of appropriate features that emphasize its role in classification. For this it consider the statistical method like correlation coefficient to identify the features from feature set whose data are very important for existing classes. The several methods such as Gaussian process, linear regression and Euclidean distance have taken into consideration for clarity of classification. The experimental results reveal that the proposed method identifies the exact relevant features for several classes.

Comparative Analysis of the Surface Roughness Parameters due to the Machining Uniformity

2016 ◽  
Vol 367 ◽  
pp. 25-33 ◽  
Author(s):  
Andrzej Golabczak ◽  
Andrzej Konstantynowicz ◽  
Marcin Golabczak
Keyword(s):  
Surface Roughness ◽  
Comparative Analysis ◽  
Magnesium Alloy ◽  
Euclidean Distance ◽  
Experimental Results ◽  
Data Organization ◽  
Roughness Parameters ◽  
Statistical Parameters ◽  
Surface Machining ◽  
3D Profile

In the paper a new method has been proposed for the determining of the very fine machining uniformity over the elaborated surface and could be applied to different machined materials and machining procedures. The proposed methodology is relatively simple and is essentially formulated in the few subsequent steps: taking surface roughness 3D profile accordingly proposed scheme; estimation of the roughness statistical parameters: Rp, Rv, Rt, Ra, Rq, Rskew, Rkurt, and if need be – surface rugosity Ru; calculation of the centroid of the obtained data due to the measurement fields, calculation of the barycentre of the obtained data with the weighting variable chosen for the appropriate evaluation of the surface machining uniformity. As the main Cartesian coordinates of the centroid calculation we propose (Rskew, Rkurt), although other data organization schemes have also been provided as the example solutions. The final evaluation of the surface machining uniformity is based upon the Euclidean distance between the centroid and barycentre of the surface roughness data. The proposed method has been applied to experimental results obtained with the AFM technique used on samples of the polished AZ31 magnesium alloy. The surface machining procedure comprised of four stages performed with using different abrasive media, finally lead to the highest grade of the surface roughness.

Sign in / Sign up

Export Citation Format

Share Document