• ETRI Journal
• /
• v.31 no.4
• /
• pp.438-447
• /
• 2009

In this paper, a novel image projection technique for face recognition application is proposed which is based on linear discriminant analysis (LDA) combined with the relevance-weighted (RW) method. The projection is performed through 2-directional and 2-dimensional LDA, or $(2D)^2$LDA, which simultaneously works in row and column directions to solve the small sample size problem. Moreover, a weighted discriminant hyperplane is used in the between-class scatter matrix, and an RW method is used in the within-class scatter matrix to weigh the information to resolve confusable data in these classes. This technique is called the relevance-weighted $(2D)^2$LDA, or RW$(2D)^2$LDA, which is used for a more accurate discriminant decision than that produced by the conventional LDA or 2DLDA. The proposed technique has been successfully tested on four face databases. Experimental results indicate that the proposed RW$(2D)^2$LDA algorithm is more computationally efficient than the conventional algorithms because it has fewer features and faster times. It can also improve performance and has a maximum recognition rate of over 97%.

• Nuclear Engineering and Technology
• /
• v.54 no.5
• /
• pp.1570-1578
• /
• 2022

During a severe accident, steam generator (SG) tubes undergo rapid changes in the pressure and temperature. Therefore, an appropriate creep model to predict a short term creep damage is essential. In this paper, a novel creep model for Alloy 690 SG tube material was proposed. It is based on the theta (θ) projection method that can represent all three stages of the creep process. The original θ projection method poses a limitation owing to its inability to represent experimental creep curves for SG tube materials for a large strain rate in the tertiary creep region. Therefore, a new modified θ projection method is proposed; subsequently, a master curve for Alloy 690 SG material is also proposed to optimize the creep model parameters, θ_i (i = 1-5). To adapt the implicit creep scheme to the finite element code, a partial derivative of incremental creep with respect to the stress is necessary. Accordingly, creep model parameters with a strictly linear relationship with the stress and temperature were proposed. The effectiveness of the model was validated using a commercial finite element analysis software. The creep model can be applied to evaluate the creep rupture behavior of SG tubes in nuclear power plants.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.842-847
• /
• 2005

This paper proposes a novel real-time EMG pattern recognition for the control of a multifunction myoelectric hand from four channel EMG signals. To cope with the nonstationary signal property of the EMG, features are extracted by wavelet packet transform. For dimensionality reduction and nonlinear mapping of the features, we also propose a linear-nonlinear feature projection composed of PCA and SOFM. The dimensionality reduction by PCA simplifies the structure of the classifier, and reduces processing time for the pattern recognition. The nonlinear mapping by SOFM transforms the PCA-reduced features to a new feature space with high class separability. Finally a multilayer neural network is employed as the pattern classifier. We implement a real-time control system for a multifunction virtual hand. From experimental results, we show that all processes, including virtual hand control, are completed within 125 msec, and the proposed method is applicable to real-time myoelectric hand control without an operation time delay.

• The korean journal of orthodontics
• /
• v.45 no.6
• /
• pp.282-288
• /
• 2015

Objective: This study investigated whether it is possible to use a two-dimensional (2D) standard in three-dimensional (3D) analysis, by comparing the angles and lengths measured from a midsagittal projection in 3D cone-beam computed tomography (CBCT) with those measured by 2D lateral cephalometric radiography (LCR). Methods: Fifty patients who underwent both LCR and CBCT were selected as subjects. CBCT was reoriented in 3 different methods and the measuring-points were projected onto the midsagittal plane. Twelve angle values and 8 length values were measured on both LCR and CBCT and compared. Results: Repeated measures analysis of the variance revealed statistically significant differences in 7 angular and 5 linear measurements among LCR and 3 types of CBCT (p < 0.05). Of these 12 measurements, multiple comparisons showed that 6 measurements (ANB, AB to FH, IMPA, FMA, Co-Gn, Go-Me) were not significantly different in pairwise comparisons. LCR was significantly different from 3 types of CBCT in 3 angular (SN to FH, interincisal angle, FMIA) and 2 linear (S-Go, Co-ANS) measurements. The CBCT method was similar for all measurements, except for 1 linear measurement, i.e., S-N. However, the disparity between the mean values for all parameters was within the range of clinical measurement error. Conclusions: 3D-CBCT analysis, using midsagittal projection, is a useful method in which the 2D-LCR normative values can be used. Although the measurements changed with reorientation, these changes were not clinically significant.

• The Korean Journal of Applied Statistics
• /
• v.25 no.2
• /
• pp.239-249
• /
• 2012

We estimate nonlinear impulse responses of stock indices by asset class by the Local Projection method as suggested by Jorda (2005) to compute impulse responses. The method estimates impulse responses without the specification and estimation of the underlying multivariate dynamic system unlike the usual way of vector autoregression(VAR). It estimates Local Projections at each period of interest rather than extrapolating into increasingly distant horizons with the advantages of easy estimation and non-linear flexible specification. The Local Projection method adequately captures the nonlinearity and asymmetry of the impulse responses of the stock indices compared to those from VARs.

• The Journal of the Acoustical Society of Korea
• /
• v.25 no.3E
• /
• pp.89-94
• /
• 2006

For audio indexing and targeted search of specific audio or corresponding visual contents, the MPEG-7 standard has adopted a sound classification framework, in which dimension-reduced Audio Spectrum Projection (ASP) features are used to train continuous hidden Markov models (HMMs) for classification of various sounds. The MPEG-7 employs Principal Component Analysis (PCA) or Independent Component Analysis (ICA) for the dimensional reduction. Other well-established techniques include Non-negative Matrix Factorization (NMF), Linear Discriminant Analysis (LDA) and Discrete Cosine Transformation (DCT). In this paper we compare the performance of different dimensional reduction methods with Gaussian mixture models (GMMs) and HMMs in the classifying video sound clips.

• Journal of information and communication convergence engineering
• /
• v.16 no.1
• /
• pp.17-22
• /
• 2018

In massive multiple-input multiple-output (MIMO) systems, linear channel estimation algorithms are widely applied owing to their simple structures. However, they may cause pilot contamination, which affects the subsequent data detection performance. Therefore, herein, for an uplink multicell massive multiuser MIMO system, we consider using an alternating projection (AP) for channel estimation to eliminate the effect of pilot contamination and improve the performance of data detection in terms of the bit error rates as well. Even though the AP is nonlinear, it iteratively searches the best solution in only one dimension, and the computational complexity is thus modest. We have analyzed the mean square error with respect to the signal-to-interference ratios for both the cooperative and non-cooperative multicell scenarios. From the simulation results, we observed that the channel estimation results via the AP benefit the following signal detection more than that via the least squares for both the cooperative and non-cooperative multicell scenarios.

• Journal of International Society for Simulation Surgery
• /
• v.2 no.2
• /
• pp.64-66
• /
• 2015

Purpose : In this paper we proposed cascade feature filter and projection method for rapid human face recognition for the large-scale high-dimensional face database. Materials and Methods : The relevant features are selected from the large feature set using Fast Correlation-Based Filter method. After feature selection, project them into discriminant using Principal Component Analysis or Linear Discriminant Analysis. Their cascade method reduces the time-complexity without significant degradation of the performance. Results : In our experiments, the ORL database and the extended Yale face database b were used for evaluation. On the ORL database, the processing time was approximately 30-times faster than typical approach with recognition rate 94.22% and on the extended Yale face database b, the processing time was approximately 300-times faster than typical approach with recognition rate 98.74 %. Conclusion : The recognition rate and time-complexity of the proposed method is suitable for real-time face recognition system on the large-scale high-dimensional face database.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.11 no.1
• /
• pp.82-87
• /
• 2012

For different ferrite-pearlite matrix structure, contain more than 90% spheroidal ratio of graphite, GCD 45-3, GCD 50, GCD 60 series and 70%, 80%, 90% spheroidal ratio of graphite, GCD 40, GCD 45-1, GCD 45-2 series, this paper has carried out rotary bending fatigue test, estimated maximum and mean size of spheroidal graphite, investigated correlation. It was concluded as follows. (1) Fatigue limit in $10^7$cycles and numbers of spheroidal graphite per 1$mm^2$ was linear relation. (2) projection area of graphite can be used to predict fatigue limit of Ductile Cast Iron. The Statistical distribution of extreme values of projection area of defects may be used as a guideline for the control of inclusion size in the steelmaking processes.

• Proceedings of the KIEE Conference
• /
• 2007.04a
• /
• pp.395-397
• /
• 2007

In this paper, a Gauss-Seidel fast affine projection (GS-FAP) algorithm developed for the linear active noise control (ANC) is further utilized for nonlinear ANC of a 2nd-order Volterra systems with a nonlinear primary path and a noisy secondary path. The simulation results, obtained by applying adaptive Volterra filtering, show that the proposed approach yields more stable and faster nonlinear AN.C, compared with the conventional methods for the nonlinear ANC in case of noisy plant models.