• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 1998.06b
• /
• pp.16.1-164
• /
• 1998

This paper describes Facial Expression Explorer to search for the components of a facial expression and to map the expression to other expressionless figures like a robot, frog, teapot, rabbit and others. In general, it is a time-consuming and laborious job to create a facial expression manually, especially when the facial expression must personify a well-known public figure or an actor. In order to extract a blending ratio from facial images automatically, the Facial Expression Explorer uses Networked Genetic Algorithm(NGA) which is a fast method for the convergence by GA. The blending ratio is often used to create facial expressions through shape blending methods by animators. With the Facial Expression Explorer a realistic facial expression can be modeled more efficiently.

• Proceedings of the KIEE Conference
• /
• 2001.07a
• /
• pp.3-5
• /
• 2001

When a fault occurs on distribution network. blackout region may happen, then it should be restored as fast as possible to minimize interruption of electric service. In this paper. A near optimal method to restore distribution network is proposed. For an optimal restoration, the number or switching operations must be minimized. The proposed method generates a general restoration plan for any distribution network fault and designed to reduce switching operations considering available load transfers. In this method overall process time can reduce with heuristic rules, which make a reduction of search space before restoration process. To achieve a near optimal solution, multiple load transfer algorithm is proposed too.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.15 no.4
• /
• pp.395-399
• /
• 2005

In this paper, we presents HIGH-Q learning algorithm with reinforcement learning for more fast and exact topic-directed web spidering. The purpose of reinforcement learning is to maximize rewards from environment, an reinforcement learning agents learn by interacting with external environment through trial and error. We performed experiments that compared the proposed method using reinforcement learning with breath first search method for searching the web pages. In result, reinforcement learning method using future discounted rewards searched a small number of pages to find result pages.

• Proceedings of the IEEK Conference
• /
• 2002.07c
• /
• pp.1539-1542
• /
• 2002

In layout of LSI and PWB, block pack- ing problem is very important in order to reduce chip area. Sequence-pair is typical one of conventional pack- ing method and can search nearly-optimal solution by using Simulated Annealing(SA). SA takes huge computation time due to evaluating of various packing results. Therefore, Sequence-pair is not effective enough for fast layout evaluation including estimation of wire length and rotation of every blocks. This paper proposes an efficient block packing method to minimize wire length and chip area. Our method searches an optimal packing efficient- ly by using a cluster growth algorithm with changing the most valuable packing score on packing process.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.36S no.5
• /
• pp.48-55
• /
• 1999

본 논문에서는 움직임 추정시 인접된 블록간의 탐색 영역이 중첩됨을 이용하여 전역 탐색 기법과 동일한 성능을 유지하면서 계산량을 현저히 감소시킬수 있는 고속 움직임 추정 기법을 제안하였다. 제안한 기법에서는 현재 움직임 추정을 행하고자 하는 블록에 대한 탐색점 중에서 인접된 블록과 중첩되는 탐색점에 대하여는 먼저 평균 절대 오차 (mean absoulte difference; MAD)를 계산할 필요가 있을지에 대한 판별을 행한 뒤 MAD 계산이 필요한 경우에 대하여서만 MAD를 구한다. MAD 계산 여부에 대한 판별에는 현재 블록과 인접 블록간의 MAD 와 인접 블록의 각 탐색점에 대한 MAD를 이용한다. 여기에 사용된 현재 블록과 인접 블록간의 MAD는 각 블록에 대하여 한번만 계산하면 되고, 인접 블록의 각 탐색점에 대한 MAD는 이미 구해져 있기 때문에 한번의 MAD 계산을 추가함으로써 탐색점 수를 현저히 감소시킬 수 있었다. 컴퓨터 모의 실험 결과로부터 제안한 방법이 전역 탐색 알고리즘과 동일한 성능을 유지하면서 많은 계산량의 감소를 얻을 수 있음을 확인 할 수 있었다.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.9 no.6
• /
• pp.649-656
• /
• 2014

In this paper, we address the problem of center lane detection algorithm for autonomous driving. Color information for center lane is gathered by analyzing a row line color distribution of road in front of a vehicle. The candidate pixels for center lane are extracted from the histogram of road colors. Morphological filtering and clustering process are applied to the candidate pixels to extract the exact center lane. We predict a expected area of center lane and search only the regions in subsequent frames, that reduces the time required for center lane detection.

• Journal of Electrical Engineering and information Science
• /
• v.2 no.5
• /
• pp.79-88
• /
• 1997

This paper introduces a new energy function, as maximum a posteriori(MAP) estimate of binocular disparity, that can deal with both random dot stereo-gram(RDS) and natural scenes. The energy function uses phase-magnitude as features to detect only the shift for a pair of corrupted conjugate images. Also we adopted Fleet singularity that effectively detects unstable areas of image plant and thus eliminates in advance error-prone stereo mathcing. The multi-scale concept is applied to the multi laser architecture that can search the solutions systematically from coarse to fine details and thereby avoids drastically the local minima. Using mean field approximation, we obtained a compact representation that is suitable for fast computation. In this manner, the energy function satisfies major natural constraints and requirements for implementing parallel relaxation. As an experiment, the proposed algorithm is applied to RDS and natural stereo images. As a result we will see that it reveals good performance in terms of recognition errors, parallel implementation, and noise characteristics.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.10b
• /
• pp.712-714
• /
• 2003

서로 다른 형태와 크기를 가지는 탐색패턴과 움직임 벡터의 분포는 블록 정합 기법에서 탐색 속도와 화질을 좌우하는 중요한 요소이다. 본 논문에서는 납작한 육각패턴을 이용한 새로운 고속 블록 정합 알고리즘을 제안한다. 이 방법은 작은 육각패턴을 이용하여 적은 탐색점으로 움직임이 적은 벡터를 우선 찾은 다음에 움직임이 큰 벡터에 대해서는 납작한 육각패턴을 이용하여 고속으로 움직임 벡터를 찾게 하였다. 실험결과, 제안된 알고리즘은 육각패턴 탐색기법에 비하여 움직임 벡터 예측의 속도에 있어서 약 11~51% 이상의 높은 성능 향상을 보였으며 화질 또한 PSNR 기준으로 약 0.05~0.74dB 의 향상을 보였다.

• Journal of Computing Science and Engineering
• /
• v.1 no.2
• /
• pp.162-176
• /
• 2007

Machine Learning (ML) has become very popular within Data Mining (KDD) and Artificial Intelligence (AI) research and their applications. In the ML and KDD contexts, two main approaches can be used for inducing a Bayesian Network (BN) from data, namely, Conditional Independence (CI) and the Heuristic Search (HS). When a BN is induced for classification purposes (Bayesian Classifier - BC), it is possible to impose some specific constraints aiming at increasing the computational efficiency. In this paper a new CI based approach to induce BCs from data is proposed and two algorithms are presented. Such approach is based on the Markov Blanket concept in order to impose some constraints and optimize the traditional PC learning algorithm. Experiments performed with the ALARM, as well as other six UCI and three artificial domains revealed that the proposed approach tends to execute fewer comparison tests than the traditional PC. The experiments also show that the proposed algorithms produce competitive classification rates when compared with both, PC and Naive Bayes.

• The Transactions of the Korea Information Processing Society
• /
• v.4 no.10
• /
• pp.2640-2650
• /
• 1997

In this paper we propose new image coding using wavelet transform. The new method constructs hierarchical bit plane and progressively transports each bit plane. The proposed algorithms not only supports multi-resolution, dividing original image into special band and various resolution but also reduces blocking effects that come int JPEG. In encoding time this algorithm considers each band characters and priority of transport order, and applies to fast search of image.