• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2006.05a
• /
• pp.1071-1078
• /
• 2006

Robust design of products or processes is a way to produce products or processes with less variability on condition that the product specification is met, so that the ratio of nonconforming products can be as minimized as possible. In such models as robust parameter design(RPD) originated by Taguchi and response surface methodology(RSM), the variances of noise variables are assumed to be measurable. However, there are circumstances in which the experimenters are unable to measure noise variables or perhaps are not aware of what the noise variables are. In this paper, it is demonstrated through an example that Taguchi's RPD approach can be conducted within the framework of the analysis of variability, with a view to a comparison of the two methods.

• Proceedings of the KIEE Conference
• /
• 2001.11c
• /
• pp.26-29
• /
• 2001

In a target tracking problem the radar glint noise has non-Gaussian heavy-tailed distribution and will seriously affect the target tracking performance. In most nonlinear situations an Extended Robust Kalman Filter(ERKF) can yield acceptable performance as long as the noises are white Gaussian. However, an Extended Robust $H_{\infty}$ Filter (ERHF) can yield acceptable performance when the noises are Laplacian. In this paper, we use the Interacting Multiple Model(IMM) estimator for the problem of target tracking with glint noise. In the IMM method, two filters(ERKF and ERHF) are used in parallel to estimate the state. Computer simulations of a real target tracking shows that hybrid filter used the IMM algorithm has superior performance than a single type filter.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.34S no.6
• /
• pp.58-66
• /
• 1997

This paper proposes a multiresolution surface parameter estimation method for range images. Based on robust estimation of surface parameters, it approximates a patch to a planar surface in the locally adaptive window. Selection of resolution is made pixelwise by comparing a locally computed homogeneity measure with th eglobal threshold determined by te distribution of the approximation error. The proposed multiresolution surface parameter estimation method is applied to range image reconstruction. Computer simulation results with noisy rnag eimages contaminated by additive gaussian noise and impulse noise show that the proposed multiresolution reconstruction method well preserves step and roof edges compared with the conventional methods. Also the segmentation method based on the estimated surface parameters is shown to be robust to noise.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.4 no.2
• /
• pp.254-258
• /
• 2004

In this paper, we propose a robust speech recognition system under noisy environments. Since the presence of noise severely degrades the performance of speech recognition system, it is important to design the robust speech recognition method against noise. The proposed method adopts a new distance measure technique based on stochastic probability instead of conventional method using minimum error. For evaluating the performance of the proposed method, we compared it with conventional distance measure for the 10-isolated Korean digits with car noise. Here, the proposed method showed better recognition rate than conventional distance measure for the various car noisy environments.

• Speech Sciences
• /
• v.10 no.3
• /
• pp.263-277
• /
• 2003

One of major problems in speech recognition is performance degradation due to the mismatch between the training and test environments. Recently, Stereo-based Piecewise LInear Compensation for Environments (SPLICE), which is frame-based bias removal algorithm for cepstral enhancement using stereo training data and noisy speech model as a mixture of Gaussians, was proposed and showed good performance in noisy environments. In this paper, we propose several methods to improve the conventional SPLICE. First we apply Cepstral Mean Subtraction (CMS) as a preprocessor to SPLICE, instead of applying it as a postprocessor. Secondly, to compensate residual distortion after SPLICE processing, two-stage SPLICE is proposed. Thirdly we employ phonetic information for training SPLICE model. According to experiments on the Aurora 2 database, proposed method outperformed the conventional SPLICE and we achieved a 50% decrease in word error rate over the Aurora baseline system.

• The Journal of the Acoustical Society of Korea
• /
• v.22 no.4E
• /
• pp.161-166
• /
• 2003

We present an adaptive wavelet packet based speech enhancement method with robust voice activity detection (VAD) in non-stationary noise environment. The proposed method can be divided into two main procedures. The first procedure is a VAD with adaptive wavelet packet transform. And the other is a speech enhancement procedure based on the proposed VAD method. The proposed VAD method shows remarkable performance even in low SNRs and non-stationary noise environment. And subjective evaluation shows that the performance of the proposed speech enhancement method with wavelet bases is better than that with Fourier basis.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.39 no.7
• /
• pp.729-739
• /
• 1990

This paper deals with the fault detection problem in uncertain linear/non-linear systems having both undermodelling and noise. A robust fault detection method is presented which accounts for the effects of noise, model mismatch and nonlinearities. The basic idea is to embed the unmodelled dynamics in a stochastic process and to use the nominal model with a predetermined fixed denominator. This allows the input /output relationship to be represented as a linear function of the system parameters and also facilitate the quatification of the effect of noise, model mismatch and linearization errors on parameter estimation by the Bayesian method. Comparisons are made via simulations with traditional fault detection methods which do not account for model mismatch or linearization errors. The new method suggested in this paper is shown to have a marked improvement over traditional methods on a number of simulations, which is a consequence of the fact that the new method explicitly for the effects of undermodelling and linearization errors.

• 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 Korean Journal of Applied Statistics
• /
• v.28 no.5
• /
• pp.991-1001
• /
• 2015

Digital images acquired by digital devices are used in many fields. Applying statistical methods to the processing of images will increase speed and efficiency. Methods to remove noise and image quality have been researched as a basic operation of image processing. This paper proposes a novel reduction method that considers the direction and magnitude of the edge to remove image noise effectively using statistical methods. The proposed method estimates the brightness of pixels relative to pixels in the same direction based on a robust regression model. An estimate of pixel brightness is obtained by weighting the magnitude of the edge that improves the performance of the average filter. As a result of the simulation study, the proposed method retains pixels that are well-characterized and confirms that noise reduction performance is improved over conventional methods.

• ETRI Journal
• /
• v.38 no.1
• /
• pp.70-80
• /
• 2016

This paper addresses the problem of robust waveform design for distributed multiple-radar systems (DMRSs) based on low probability of intercept (LPI), where signal-to-interference-plus-noise ratio (SINR) and mutual information (MI) are utilized as the metrics for target detection and information extraction, respectively. Recognizing that a precise characterization of a target spectrum is impossible to capture in practice, we consider that a target spectrum lies in an uncertainty class bounded by known upper and lower bounds. Based on this model, robust waveform design approaches for the DMRS are developed based on LPI-SINR and LPI-MI criteria, where the total transmitting energy is minimized for a given system performance. Numerical results show the effectiveness of the proposed approaches.