• ETRI Journal
• /
• v.32 no.5
• /
• pp.761-765
• /
• 2010

In this paper, we introduce an improved adaptive boosting (AdaBoost) classifier and its application, a disguised-face discriminator that discriminates between bare and disguised faces. The proposed classifier is based on an AdaBoost learning algorithm and regression technique. In the process, the lookup table of AdaBoost learning is utilized. The proposed method is verified on the captured images under several real environments. Experimental results and analysis show the proposed method has a higher and faster performance than other well-known methods.

• ETRI Journal
• /
• v.28 no.4
• /
• pp.533-536
• /
• 2006

A distributed Bragg reflector (DBR) laser and a high speed electroabsorption modulator (EAM) are integrated on the basis of the selective area growth technique. The typical threshold current is 4 to 6 mA, and the side mode suppression ratio is over 40 dB with single mode operation at 1550 nm. The DBR laser exhibits 2.5 to 3.3 mW fiber output power at a laser gain current of 100 mA, and a modulator bias voltage of 0 V. The 3 dB bandwidth is 13 GHz. A 10 Gbps non-return to zero operation with 12 dB extinction ratio is obtained. A four-channel laser array with 100 GHz wavelength spacing was fabricated and its operation at the designed wavelength was confirmed.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.16 no.11
• /
• pp.1194-1200
• /
• 1991

The FEM is a computer-aided mathematical technique for obtaining approximate solution to the differential equations. The pointwise convergence defines the relationship between the mesh size and the tolerance. This will play an important role in improving quality of finite element approximate solution. In the paper. We evaluate the convergence on a certain unknown point with a 1-irregular mesh refinement and spectral order enrichment. This means that the degree of freedom is minimized within a tolerance.

• International Journal of Control, Automation, and Systems
• /
• v.6 no.2
• /
• pp.194-203
• /
• 2008

Iterative learning control (ILC) is a simple and effective method for the control of systems that perform the same task repetitively. ILC algorithm uses the repetitiveness of the task to track the desired trajectory. In this paper, we propose a PID (proportional plus integral and derivative) type ILC update law for control discrete-time single input single-output (SISO) linear time-invariant (LTI) systems, performing repetitive tasks. In this approach, the input of controlled system in current cycle is modified by applying the PID strategy on the error achieved between the system output and the desired trajectory in a last previous iteration. The convergence of the presented scheme is analyzed and its convergence condition is obtained in terms of the PID coefficients. An optimal design method is proposed to determine the PID coefficients. It is also shown that under some given conditions, this optimal iterative learning controller can guarantee the monotonic convergence. An illustrative example is given to demonstrate the effectiveness of the proposed technique.

• Journal of information and communication convergence engineering
• /
• v.2 no.3
• /
• pp.177-181
• /
• 2004

In this paper, we suggest an Iris recognition system with an excellent recognition rate and confidence as an alternative biometric recognition technique that solves the limit in an existing individual discrimination. For its implementation, we extracted coefficients feature values with the wavelet transformation mainly used in the signal processing, and we used neural network to see a recognition rate. However, Scale Conjugate Gradient of nonlinear optimum method mainly used in neural network is not suitable to solve the optimum problem for its slow velocity of convergence. So we intended to enhance the recognition rate by using Levenberg-Marquardt Back-propagation which supplements existing Scale Conjugate Gradient for an implementation of the iris recognition system. We improved convergence velocity, efficiency, and stability by changing properly the size according to both convergence rate of solution and variation rate of variable vector with the implementation of an applied algorithm.

• International journal of advanced smart convergence
• /
• v.8 no.2
• /
• pp.183-190
• /
• 2019

In the era of the Fourth Industrial Revolution, digital transformation, which means changes in all industrial structures, politics, economics and society as well as IT technology, is an important issue. It is difficult to know which research topic is being studied because digital transformation is being studied in various fields. Convergence research is possible because a research topic is studied in various fields such as computer science area and Decision science area. However, it is difficult to know the specific research status of the research topic. In this study, eight research topics were derived using the topic modeling technique of text mining for abstract of academic literature and the trend of each topic was analyzed. We also proposed to create a Topic-Word Proportions Table in the LDA based Topic modeling process to predict the topic of new literature. The results of this study are expected to contribute to advanced convergence research on topic of digital transformation. It is expected that the literature related to each research topic will be grasped and contribute to the design of a new convergence research.

• Journal of Information Processing Systems
• /
• v.19 no.1
• /
• pp.33-45
• /
• 2023

This paper proposes an algorithm that automatically converts images into Pipo, painting images using OpenCV-based image processing technology. The existing "purity," "palm," "puzzling," and "painting," or Pipo, painting image production method relies on manual work, so customized production has the disadvantage of coming with a high price and a long production period. To resolve this problem, using the OpenCV library, we developed a technique that automatically converts an image into a Pipo painting image by designing a module that changes an image, like a picture; draws a line based on a sector boundary; and writes sector numbers inside the line. Through this, it is expected that the production cost of customized Pipo painting images will be lowered and that the production period will be shortened.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.07a
• /
• pp.193-194
• /
• 2021

영화, 드라마 등과 같은 콘텐츠에서 표현되는 감정은 등장인물의 대화와 표정뿐만이 아니라, 영상이 표현하는 다양한 정보 중 하나인 촬영기법, 장면의 배경 등을 통해서도 표현된다. 특히 핸드헬드 샷은 불안정하지만 현장감과 자유분방한 감정을 관객에게 전달하며 긴장감, 공포 등 배우들의 감정선을 따라가게 하는 효과가 있다. 따라서 영상 콘텐츠에서 감정 정보를 분석하기 위해서는 핸드헬드 샷을 검출하는 것은 기초적인 작업에 해당한다. 본 논문에서는 핸드헬드 샷을 양방향 LSTM을 활용하여 구별하는 방법을 제안한다. 제안된 방법으로 인식한 핸드헬드의 인식 정확도는 97%였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2024.05a
• /
• pp.171-174
• /
• 2024

IoT 가 전 산업에 널리 활용되면서 생성되는 데이터 양이 급증하고 있다. 그러나 경량, 저가, 저전력 IoT 는 대용량 데이터를 처리, 저장, 전송하기 어렵다. 그러나 이러한 문제를 해결하기 위한 종래의 방법들은 복잡도와 성능의 트레이드오프 문제가 있다. 본 논문은 IoT 기기의 효율적 리소스 사용을 위한 오토 인코더 데이터 압축 기법을 제안한다. 실험 결과에 따르면 제안한 기법은 종래 기술에 비해 평균 60.61% 축소된 데이터 크기를 보였다. 또한, 제안된 기법으로 압축된 데이터를 사용하여 모델 학습을 진행한 결과에 따르면 RNN 과 LSTM 모델에 제안한 방법을 적용했을 때 모두 97% 이상의 정확도를 보였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2024.05a
• /
• pp.290-293
• /
• 2024

최근 IoT (Internet of Things, IoT) 기기가 전 산업과 일상 생활에 활용되면서 취약점 탐지 기술이 중요해지고 있다. 그러나 리소스가 제약적인 IoT 기기에는 종래의 퍼징 기술을 적용하기 어렵다. 본 논문에서는 경량화 IoT 환경에서 퍼징 기술을 적용하기 위한 테스트케이스 선택 기법을 제안했다. 실험 결과에 따르면, 제안하는 방식은 무작위 입력을 생성하여 퍼징하는 종래 퍼저보다 평균 61.49% 빠르게 취약점을 탐지했다.