• The Journal of the Acoustical Society of Korea
• /
• v.15 no.5
• /
• pp.38-43
• /
• 1996

This paper compares the recognition performances of feature vectors known to be robust to the environmental noise. And, the speech subtraction technique is combined with the noise robust feature to get more performance enhancement. The experiments using SMC(Short time Modified Coherence) analysis, root cepstral analysis, LDA(Linear Discriminant Analysis), PLP(Perceptual Linear Prediction), RASTA(RelAtive SpecTrAl) processing are carried out. An isolated word recognition system is composed using semi-continuous HMM. Noisy environment experiments usign two types of noises:exhibition hall, computer room are carried out at 0, 10, 20dB SNRs. The experimental result shows that SMC and root based mel cepstrum(root_mel cepstrum) show 9.86% and 12.68% recognition enhancement at 10dB in compare to the LPCC(Linear Prediction Cepstral Coefficient). And when combined with spectral subtraction, mel cepstrum and root_mel cepstrum show 16.7% and 8.4% enhanced recognition rate of 94.91% and 94.28% at 10dB.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.9 no.4
• /
• pp.897-904
• /
• 2005

Face Recognition has been an active research area because it is not difficult to acquire face image data and it is applicable in wide range area in real world. Due to the high dimensionality of a face image space, however, it is not easy to process the face images. In this paper, we propose a method to reduce the dimension of the facial data and extract the features from them. It will be solved using the method which extracts the features from holistic face images. The proposed algorithm consists of two parts. The first is the using of principal component analysis (PCA) to transform three dimensional color facial images to one dimensional gray facial images. The second is integrated linear discriminant analusis (PCA+LDA) to prevent the loss of informations in case of performing separated steps. Integrated LDA is integrated algorithm of PCA for reduction of dimension and LDA for discrimination of facial vectors. First, in case of transformation from color image to gray image, PCA(Principal Component Analysis) is performed to enhance the image contrast to raise the recognition rate. Second, integrated LDA(Linear Discriminant Analysis) combines the two steps, namely PCA for dimensionality reduction and LDA for discrimination. It makes possible to describe concise algorithm expression and to prevent the information loss in separate steps. To validate the proposed method, the algorithm is implemented and tested on well controlled face databases.

• Economic and Environmental Geology
• /
• v.38 no.1
• /
• pp.45-55
• /
• 2005

Traditional GIS-based probabilistic spatial data integration models for landslide susceptibility analysis have failed to provide the theoretical backgrounds and effective methods for integration of different types of spatial data such as categorical and continuous data. This paper applies two spatial data integration models including non-parametric empirical estimation and parametric predictive discriminant analysis models that can directly use the original continuous data within a likelihood ratio framework. Similarity rates and a prediction rate curve are computed to quantitatively compare those two models. To illustrate the proposed models, two case studies from the Jangheung and Boeun areas were carried out and analyzed. As a result of the Jangheung case study, two models showed similar prediction capabilities. On the other hand, in the Boeun area, the parametric predictive discriminant analysis model showed the better prediction capability than that from the non-parametric empirical estimation model. In conclusion, the proposed models could effectively integrate the continuous data for landslide susceptibility analysis and more case studies should be carried out to support the results from the case studies, since each model has a distinctive feature in continuous data representation.

• Journal of the Institute of Convergence Signal Processing
• /
• v.4 no.2
• /
• pp.18-24
• /
• 2003

One of the most important subjects in the implementation of real time speech recognizer is to design both reliable VAD(Voice Activity Detection) and suitable speech feature vector. But, because it is difficult to calculate reliable VAD in the environment having surrounding noise, designed suitable speech feature vector may not be obtained. Solving this problem, in this paper, we implement not only short time power spectrum which is generally used but also two additive parameters, the comparison measure of spectrum density having robust property in noise and linear discriminant function using linear regression, then perform VAD by using the combination of each parameter having apt weight in other magnitudes of surrounding noise and confirm that proposed parameters show a robust characteristic in circumstances having surrounding noise by using DTW(Dynamic Time Waning) in recognition experiment.

• The KIPS Transactions:PartB
• /
• v.17B no.1
• /
• pp.37-46
• /
• 2010

In this paper we have presented a two dimensional model based tracking system using improved chamfer matching. Conventional chamfer matching could not calculate similarity well between the object and image when there is very cluttered background. Then we have improved chamfer matching to calculate similarity well even in very cluttered background with edge and corner feature points. Improved chamfer matching is used as likelihood function of particle filter which tracks the geometric object. Geometric model which uses edge and corner feature points, is a discriminant descriptor in color changes. Particle Filter is more non-linear tracking system than Kalman Filter. Then the presented method uses geometric model, particle filter and improved chamfer matching for tracking object in complex environment. In experimental result, the robustness of our system is proved by comparing other methods.

• Journal of Korea Multimedia Society
• /
• v.15 no.5
• /
• pp.606-614
• /
• 2012

Face detection using AdaBoost algorithm is one of the fastest and the most robust face detection algorithm so many improvements or extensions of this method have been proposed. However, almost all previous approaches deal with only frontal face and suffer from limited discriminant capability for inclined face because these methods apply the same features for both frontal and inclined face. Also conventional approaches for detecting inclined face which apply frontal face detecting method to inclined input image or make different detectors for each angle require heavy computational complexity and show low detection rate. In order to overcome this problem, a method for detecting inclined face using JointBoost is proposed in this paper. The computational and sample complexity is reduced by finding common features that can be shared across the classes. Simulation results show that the detection rate of the proposed method is at least 2% higher than that of the conventional AdaBoost method under the learning condition with the same iteration number. Also the proposed method not only detects the existence of a face but also gives information about the inclined direction of the detected face.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.13 no.6
• /
• pp.1333-1342
• /
• 2018

Until now, Electroencephalography(: EEG) has been the most important and convenient method for the diagnosis and treatment of epilepsy. However, it is difficult to identify the wave characteristics of an epileptic EEG signals because it is very weak, non-stationary and has strong background noise. In this paper, we analyse the effect of dimensionality reduction methods on Epileptic EEG feature selection and classification. Three dimensionality reduction methods: Pincipal Component Analysis(: PCA), Kernel Principal Component Analysis(: KPCA) and Linear Discriminant Analysis(: LDA） were investigated. The performance of each method was evaluated by using Support Vector Machine SVM, Logistic Regression(: LR), K-Nearestneighbor(: K-NN), Decision Tree(: DR) and Random Forest(: RF). From the experimental result, PCA recorded 75% of highest accuracy in SVM, LR and K-NN. KPCA recorded 85% of best performance in SVM and K-KNN while LDA achieved 100% accuracy in K-NN. Thus, LDA dimensionality reduction is found to provide the best classification result for epileptic EEG signal.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.5
• /
• pp.197-206
• /
• 2014

Motor imagery based Brain-computer Interface(BCI), which has recently attracted attention, is the technique for decoding the user's voluntary motor intention using Electroencephalography(EEG). For classifying the motor imagery, event-related desynchronization(ERD), which is the phenomenon of EEG voltage drop at sensorimotor area in ${\mu}$-band(8-13Hz), has been generally used but this method are not free from the performance degradation of the BCI system because EEG has low spatial resolution and shows different ERD-appearing band according to users. Common spatial pattern(CSP) was proposed to solve the low spatial resolution problem but it has a disadvantage of being very sensitive to frequency-band selection. Discriminative filter bank common spatial pattern(DFBCSP) tried to solve the frequency-band selection problem by using the Fisher ratio of the averaged EEG signal power and establishing discriminative filter bank(DFB) which only includes the feature frequency-band. However, we found that DFB might not include the proper filters showing the spatial pattern of ERD. To solve this problem, we apply a band-selection process using CSP feature vectors and linear discriminant analysis to DFBCSP instead of the averaged EEG signal power. The filter selection results and the classification accuracies of the existing and the proposed methods show that the CSP feature is more effective than signal power feature.

• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.568-571
• /
• 1997

Recently, the number of techniques for analyzing medical images has been increasing in computer vision, employing X-ray CT images, ultrasound images, MR images, moire topographic images, etc. Spinal deformity is a serious problem especially for teenagers and medical doctors inspect moire topographic images of their backs visually for the primary screening. If a subject is normal, the moire image is almost symmetric with respect to the middle line of the subject's back, otherwise it shows asymmetric shape. In this paper, an image analysis technique is described for discriminating suspicious cases from normal in human spinal deformity by recognizing asymmetric moire images of human backs. The principal axes which are sensitive to asymmetry of the moire image are extracted at two parts on a subject's back and their angles are evaluated with respect to the detected middle line of the back. The two angles compose a 2-D feature space and inspected cases are divided into two clusters in the space by a linear discriminant function based on the Mahalanobis distance. Given 120 cases, 60 normal and 60 abnormal, the leave-out method was applied for the recognition and 75% recognition rate was achieved.

• Journal of Korea Multimedia Society
• /
• v.13 no.12
• /
• pp.1786-1797
• /
• 2010

Automatic understanding of contents in document image is a very hard problem due to involvement with mathematically challenging problems originated mainly from the over-determined system induced by document segmentation process. In both academic and industrial areas, there have been incessant and various efforts to improve core parts of content retrieval technologies by the means of separating out segmentation related issues using semi-structured document, e.g., invoice,. In this paper we proposed classification models for text lines on invoice document in which text lines were clustered into the five categories in accordance with their contents: purchase order header, invoice header, summary header, surcharge header, purchase items. Our investigation was concentrated on the performance of machine learning based models in aspect of linear-discriminant-analysis (LDA) and non-LDA (logic based). In the group of LDA, na$\"{\i}$ve baysian, k-nearest neighbor, and SVM were used, in the group of non LDA, decision tree, random forest, and boost were used. We described the details of feature vector construction and the selection processes of the model and the parameter including training and validation. We also presented the experimental results of comparison on training/classification error levels for the models employed.