• Journal of KIISE:Software and Applications
• /
• v.33 no.5
• /
• pp.481-488
• /
• 2006

The performance of face recognition methods using subspace projection is directly related to the characteristics of their basis images, especially in the cases of local distortion or partial occlusion. In order for a subspace projection method to be robust to local distortion and partial occlusion, the basis images generated by the method should exhibit a part-based local representation. We propose an effective part-based local representation method named locally salient ICA (LS-ICA) method for face recognition that is robust to local distortion and partial occlusion. The LS-ICA method only employs locally salient information from important facial parts in order to maximize the benefit of applying the idea of 'recognition by parts.' It creates part-based local basis images by imposing additional localization constraint in the process of computing ICA architecture I basis images. We have contrasted the LS-ICA method with other part-based representations such as LNMF (Localized Non-negative Matrix Factorization) and LFA (Local Feature Analysis). Experimental results show that the LS-ICA method performs better than PCA, ICA architecture I, ICA architectureII, LFA, and LNMF methods, especially in the cases of partial occlusions and local distortions.

• Proceedings of the IEEK Conference
• /
• 2000.11d
• /
• pp.77-80
• /
• 2000

In this paper, we present a new automatic thresholding algorithm based on maximum entropy of two-dimensional pixel histogram. While most of the previous algorithms select thresholds depending only on the histogram of gray level itself in the image, the presented algorithm considers 2D relational histogram of gray levels of two adjacent pixels in the image. Thus, the new algorithm tends to leave salient edge features on the image after thresholding. The experimental results show the good performance of the presented algorithm.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2001.01a
• /
• pp.298-309
• /
• 2001

This research aims to unravel the significant features of the human immune system, which would be successfully employed for a novel network intrusion detection model. Several salient features of the human immune system, which detects intruding pathogens, are carefully studied and the possibility and the advantages of adopting these features for network intrusion detection are reviewed and assessed.

• Journal of Chest Surgery
• /
• v.15 no.3
• /
• pp.338-345
• /
• 1982

An 11 year old boy with a preductal coarctation and a patent ductus arteriosus underwent ligation of the ductus and patch aortoplasty for correction of the coarctation. His postoperative course was complicated by a paradoxical hypertension and a transient mesenteric arteritis syndrome which were successfully managed with sodium nitroprusside, propranolol and hydralazine. His urine catecholamine tests were markedly elevated one week after the operation which returned to normal 4 days later. Some of the salient points ~n the pathogenesis and management of these problems are discussed.

• Korean Management Science Review
• /
• v.16 no.1
• /
• pp.101-113
• /
• 1999

The individual behavior on considering prior information when one assesses the probability of uncertain event by representativeness heuristic has been investigated. While prior researches proposed two contrasting behaviors on the employment, we tested the mixed hypothesis that individual ignores the prior information to some extents and begins to consider it above certain threshold when the evidence of representativeness is not salient. The threshold effect of prior probability is positively experimented and the results strongly support the discontinuity hypothesis of representativeness heuristic.

• Proceedings of the KIEE Conference
• /
• 2005.11b
• /
• pp.217-219
• /
• 2005

This paper describes implicit restated arnoldi method algorithm and Its application to small size power systems In order to observe the salient features of IRAM algorithm. Two area system with 36 state variables and England 39-bus system with 150 state variables have been tested using IRAM, and the eigenvalue results of IRAM are compared with those of the results obtained from QR method.

• Communications for Statistical Applications and Methods
• /
• v.8 no.3
• /
• pp.867-873
• /
• 2001

In this paper we propose a family of skew- f distributions. The family is derived by a scale mixtures of skew-normal distributions introduced by Azzalini (1985) and Henze (1986). The salient features of the family are mathematical tractability and strict inclusion of the normal law. Further it includes a shape parameter, to some extent, controls the index of skewness. Necessary theory involved in deriving the family of distributions is provided and main properties of the family are also studied.

• Proceedings of the Korea Concrete Institute Conference
• /
• 2008.11a
• /
• pp.843-846
• /
• 2008

Rationality of strength models for structural concrete depends on how to treat loads on boundaries and load paths within members. Differentiation between strut-and-tie models and stress fields approaches for shear strength models is discussed in this paper for salient use of current design formula in design code provisions. How to model configuration of loads and stress states along the boundary for the regions under provides a key to realistic construction of stress fields together with STM.

• 한국정보디스플레이학회:학술대회논문집
• /
• 2008.10a
• /
• pp.1065-1066
• /
• 2008

Printed thin-film transistors (TFTs) have received profound interests as an alternative to their silicon counterparts for use in fabricating next-Gen microelectronics by virtue of projected low manufacturing cost and certain salient features (e.g., thin and lightweight characteristics, structural flexibility, etc.) that printed TFTs bring to device architecture. The economic advantages stem from engaging low-cost printing techniques (e.g., screen printing, gravure, flexography, etc.) for deposition and patterning in place of traditionally costly high-vacuum, high-temperature photolithographic processes. To render printing TFTs possible, solution processable materials are necessary.

• Proceedings of the KIPE Conference
• /
• 1998.11a
• /
• pp.1-4
• /
• 1998

In this paper, an auto-tuning method for fuzzy controller based on the neural network is presented. The backpropagated error of neural emulator offers the path which reforms the fuzzy controller's membership functions and fuzzy rule, and used for speed control of switched reluctance motor. The experiments are performed to verify the capability of proposed control method on 6/4 salient type SRM. The results show that fuzzy-neural controller is suitable for wide speed range.