• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.6
• /
• pp.89-97
• /
• 2004

An object to reflect or emit light is captured by imaging system as distorted image due to various distortion. It is called image restoration that estimates original object by removing distortion. There are two categories in image restoration method. One is a deterministic method and the other is a stochastic method. In this paper, image restoration using Minimum Fisher Information(MFI), derived from B. Roy Frieden is proposed. In MFI restoration, experimental results to be made according to noise control parameter were investigated. And cross entropy(Kullback-Leibler entropy) was used as a standard measure of restoration accuracy, It is confirmed that restoration results using MFI have various roughness according to noise control parameter.

• The Journal of the Acoustical Society of Korea
• /
• v.29 no.3
• /
• pp.216-221
• /
• 2010

In this paper, we propose speaker verification method using Support Vector Machine (SVM) kernel with Gaussian Mixture Model (GMM)-supervector based on the Mahalanobis distance. The proposed GMM-supervector SVM kernel method is combined GMM with SVM. The GMM-supervectors are generated by GMM parameters of speaker and other speaker utterances. A speaker verification threshold of GMM-supervectors is decided by SVM kernel based on Mahalanobis distance to improve speaker verification accuracy. The experimental results for text-independent speaker verification using 20 speakers demonstrates the performance of the proposed method compared to GMM, SVM, GMM-supervector SVM kernel based on Kullback-Leibler (KL) divergence, and GMM-supervector SVM kernel based on Bhattacharyya distance.

• Structural Engineering and Mechanics
• /
• v.48 no.3
• /
• pp.291-308
• /
• 2013

Structural health monitoring is crucial to maintain the structural performance safely. Moreover, the Kullback-Leibler divergence (KLD) is applied usually to asset the similarity between different probability density functions in the pattern recognition. In this study, the KLD is employed to detect the damage. However the asymmetry of the KLD is a shortcoming for the damage detection, to overcoming this shortcoming, two other divergences and one statistic distribution are proposed. Then the damage identification by the KLD and its three descriptions from the symmetric point of view is investigated. In order to improve the reliability and accuracy of the four divergences, the gapped smoothing method (GSM) is adopted. On the basis of the damage index approach, the new damage index (DI) for detect damage more accurately based on the four divergences is developed. In the last, the grey relational coefficient and hypothesis test (GRCHT) is utilized to obtain the more precise damage identification results. Finally, a clear remarkable improvement can be observed. To demonstrate the feasibility and accuracy of the proposed method, examples of an isotropic beam with different damage scenarios are employed so as to check the present approaches numerically. The final results show that the developed approach successfully located the damaged region in all cases effect and accurately.

• Nuclear Engineering and Technology
• /
• v.55 no.3
• /
• pp.827-838
• /
• 2023

Centrifugal pump is a key part of nuclear power plant systems, and its health status is critical to the safety and reliability of nuclear power plants. Therefore, fault diagnosis is required for centrifugal pump. Traditional fault diagnosis methods have difficulty extracting fault features from nonlinear and non-stationary signals, resulting in low diagnostic accuracy. In this paper, a new fault diagnosis method is proposed based on the improved particle swarm optimization (IPSO) algorithm-based variational modal decomposition (VMD) and relevance vector machine (RVM). Firstly, a simulation test bench for rotor faults is built, in which vibration displacement signals of the rotor are also collected by eddy current sensors. Then, the improved particle swarm algorithm is used to optimize the VMD to achieve adaptive decomposition of vibration displacement signals. Meanwhile, a screening criterion based on the minimum Kullback-Leibler (K-L) divergence value is established to extract the primary intrinsic modal function (IMF) component. Eventually, the factors are obtained from the primary IMF component to form a fault feature vector, and fault patterns are recognized using the RVM model. The results show that the extraction of the fault information and fault diagnosis classification have been improved, and the average accuracy could reach 97.87%.

• MALSORI
• /
• no.56
• /
• pp.225-235
• /
• 2005

Speech unit concatenation with a large database is presently the most popular method for speech synthesis. In this approach, the mismatches at the unit boundaries are unavoidable and become one of the reasons for quality degradation. This paper proposes an algorithm to reduce undesired discontinuities between the subsequent units. Optimal matching points are calculated in two steps. Firstly, the fullback-Leibler distance measurement is utilized for the spectral matching, then the unit sliding and the overlap windowing are used for the waveform matching. The proposed algorithm is implemented for the corpus-based unit concatenating Korean text-to-speech system that has an automatically labeled database. Experimental results show that our algorithm is fairly better than the raw concatenation or the overlap smoothing method.

• Communications for Statistical Applications and Methods
• /
• v.12 no.3
• /
• pp.743-758
• /
• 2005

This paper consider penalized likelihood regression with data from exponential family. The fast computation method applied to Gaussian data(Kim and Gu, 2004) is extended to non Gaussian data through asymptotically efficient low dimensional approximations and corresponding algorithm is proposed. Also smoothing parameter selection is explored for various exponential families, which extends the existing cross validation method of Xiang and Wahba evaluated only with Bernoulli data.

• Proceedings of the IEEK Conference
• /
• 2008.06a
• /
• pp.915-916
• /
• 2008

We propose a clustering based object feature matching for identification of same object in multi-camera system. The method is focused on ease to system initialization and extension. Clustering is used to estimate parameters of Gaussian mixture models of objects. A similarity measure between models are determined by Kullback-Leibler divergence. This method can be applied to occlusion problem in tracking.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.20 no.4
• /
• pp.833-839
• /
• 2016

In this paper, we propose a selective feature extraction algorithm between Markov transition probability and co-occurrence probability for an effective image splicing detection. The Features used in our method are composed of the difference values between DCT coefficients in the adjacent blocks and the value of Kullback-Leibler divergence(KLD) is calculated to evaluate the differences between the distribution of original image features and spliced image features. KLD value is an efficient measure for selecting Markov feature or Co-occurrence feature because KLD shows non-similarity of the two distributions. After training the extracted feature vectors using the SVM classifier, we determine whether the presence of the image splicing forgery. To verify our algorithm we used grid search and 6-folds cross-validation. Based on the experimental results it shows that the proposed method has good detection performance with a limited number of features compared to conventional methods.

• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.6
• /
• pp.405-413
• /
• 2009

Tagging system is the system that allows internet users to assign new meta-data which is called tag to article, photo, video and etc. for facilitating searching and browsing of web contents. Tag cloud, a visual interface is widely used for browsing tag space. Tag cloud selects the tags with the highest frequency and presents them alphabetically with font size reflecting their popularity. However the conventional tag selection method includes known weaknesses. So, we propose a novel tag selection method Freshness, which helps to find fresh web contents. Freshness is the mean value of Kullback-Leibler divergences between each consecutive change of tag co-occurrence probability distribution. We collected tag data from three web sites, Allblog, Eolin and Technorati and constructed the system, 'Fresh Tag Cloud' which collects tag data and creates our tag cloud. Comparing the experimental results between Fresh Tag Cloud and the conventional one with data from Allblog, our one shows 87.5% less overlapping average, which means Fresh Tag Cloud outperforms the conventional tag cloud.

• The Journal of the Acoustical Society of Korea
• /
• v.35 no.2
• /
• pp.83-91
• /
• 2016

Music-similarity computation is crucial in developing music information retrieval systems for browsing and classification. This paper overviews the recently-proposed centroid-model based music retrieval method and applies the distributional similarity measures to the model for retrieval-performance evaluation. Probabilistic distance measures (also called divergence) compute the distance between two probability distributions in a certain sense. In this paper, we consider the alpha divergence in computing distance between two centroid models for music retrieval. The alpha divergence includes the widely-used Kullback-Leibler divergence and Bhattacharyya distance depending on the values of alpha. Experiments were conducted on both genre and singer datasets. We compare the music-retrieval performance of the distributional similarity with that of the vector distances. The experimental results show that the alpha divergence improves the performance of the centroid-model based music retrieval.