• Annual Conference on Human and Language Technology
• /
• 1989.10a
• /
• pp.137-141
• /
• 1989

본 논문에서는 신경망 구조의 한 모델인 feed-forward multi-layered network에 역전파 학습(back-propagation learning) 기법을 이용하여 필기체 한글 숫자를 인식하고 그 가능성을 보였다. 문자 인식에 있어 입력 대상의 모양이 왜곡되거나, 대상의 크기 혹은 위치의 변화 등과 같은 잡음 (noise)에 대해서 정확히 대상을 인식하는 데는 대상의 구조 추출에 크게 관여되므로 한글의 구조 추출에 적합하다고 생각되는 bar mask 투사법을 제안하였다. 모델의 학습을 필기체 한글 숫자 16자의 입력 패턴과 타겟 ( target) 입력의 쌍을 이용해 학습시켰다. 또한, 모델의 인식 정도를 측정해 보기 위해 시험패턴을 적용하여 훈련된 패턴과 훈련되지 않은 패턴간의 인식률을 비교하여 보았다.

• Journal of the Institute of Convergence Signal Processing
• /
• v.1 no.1
• /
• pp.42-48
• /
• 2000

In this paper, we propose the face image recognition algorithm using skin color information, face region features such as eye, nose, and mouse, etc., and geometrical features of chin line. In the proposed algorithm, we used the intensity as well as skin color information in the HSI color coordinate which is similar to human eye system. The experimental results of proposed method shows improved extraction quality of face and provides adaptive extraction methods for the races. And also, we used chin line information as well as geometrical features of face such as eye, nose, mouse information for the improvement of face recognition quality, Experimental results shows the more improved recognition as well as extraction quality than conventional methods.

• Journal of the Korean Society of Hazard Mitigation
• /
• v.9 no.5
• /
• pp.9-13
• /
• 2009

Modal parameters including resonant frequencies and mode shapes are heavily utililized in most damage identification throries for structural health monitoring. However, extracting modal parameters from dynamic responses needs postprocessing which inevitably involves errors in curve-fitting resonants as well as transforming the domain of responses. In this paper, the applicability of a damage identification method based on free vibration responses to the in-sevice responses is experimentally verified. The experiment is performed via applying periodic and nonperiodic moving loads to a simply supported beam and displacement responses are measured. The moving load is simulated using steel balls and a downhill device. The damage identification results show that the in-service response may be applicable to identifying damage in the beam.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.200-203
• /
• 2021

This paper introduces a smart mirror technology that recognizes a person's facial expressions through image classification among several artificial intelligence technologies and presents them in a mirror. 5 types of facial expression images are trained through artificial intelligence. When someone looks at the smart mirror, the mirror recognizes my expression and shows the recognized result in the mirror. The dataset fer2013 provided by kaggle used the faces of several people to be separated by facial expressions. For image classification, the network structure is trained using convolution neural network (CNN). The face is recognized and presented on the screen in the smart mirror with the embedded board such as Raspberry Pi4.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2000.10a
• /
• pp.367-370
• /
• 2000

본 논문에서는 빠른 검색 시간을 가지면서 단어의 역문자열도 검색할 수 있는 사전 구조를 설계하고 구현한다. 빠른 검색 시간을 지원하고, 역문자열 검색을 효율적으로 하기 위해 트라이 구조를 이용하였으며 트라이 성질 잘 표현하는 배열을 이용한 구현 방법을 사용하였다. 이 사전 구조는 형태소분석, 정보검색, 음성인식 및 문자 인식 과정 등 다양한 분야에서 유용하게 이용할 수 있다.

• 한국ITS학회:학술대회논문집
• /
• 2004.11a
• /
• pp.208-211
• /
• 2004

본 연구는 필기체 한글 문자를 인식하기 위하여 한글 문자구조를 6개 기본구조로 분류한다. 각각의 한글 자음과 모음을 7-세크먼트, '/'(Left-Incline), '$\backslash$'(Right-Incline), '-'(Left-Right), '$\mid$'(UP-Down), 'c'(Circle), 'ㄱ'(Right-down), 'ㄴ'(Down-Right) 분리한다. 분리된 7-세크먼트에 대해 한글이 쓰여지는 위치에 따라 8개의 기본구조로 정의하여 세크먼트를 분리하여 레벨화한다. 따라서 본 연구는 문자를 자소(자음과모음)로 하여 7-세크먼트로 분리하는 필기체 자소분리 구조를 제시한다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2017.05a
• /
• pp.708-711
• /
• 2017

The purpose of this study was to examine the perception of abortion and suicide for 115 paramedic students in 2016. The questionnaire consisted of perception of abortion(8items), the perception of suicide(58items), related 6items of the abortion and related 2 items of the suicide. The collected data were analyzed through t-test, Chi-squared test and ANOVA, using SPSS 21.0. As a result, the total perception score against abortion was $20.57{\pm}4.12$ and the total score of suicide perception was $-14.16{\pm}15.76$. The most answer was that pregnant women were abortion decision makers in both male and female students, and abortions were the highest when they might be unwanted pregnancies. The group who had thought about suicide was higher of perceptions of suicide than the group who did not. The female students who had suicide attempt or suicide experience of close acquaintances showed higher suicide perception score than those who did not. Therefore, A bio-ethics education curriculum should be prepared to establish ethics that respects life for future paramedic students who must make decisions at the boundary between death and life. It is necessary to raise the level of positive suicide perceptions through continuing counseling and education focusing on students who have experienced suicide by family members or acquaintances.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.14 no.3
• /
• pp.279-284
• /
• 2004

Fingerprints have been widely used in the biometric authentication because of its performance, uniqueness and universality. Recently, the speed of identification becomes a very important point in the fingerprint-based security applications. Also, the reliability still remains the main issue in the fingerprint identification. In this paper, we propose the fast and reliable fingerprint matching algorithm based on the process of the 'self-nonself' discrimination in the biological immune system. The proposed algorithm is organized by two-matching stage. The 1st matching stage does the matching process by the use of the 'self-space' and MHC detector string set that are generated from the minutiae and the values of the directional field. Then the 2nd matching stage is made based on the local-structure of the minutiae. The proposed two matching stage reduces matching time while the reliability of the matching algorithm is maintained.

• The Journal of the Korea Contents Association
• /
• v.18 no.3
• /
• pp.148-159
• /
• 2018

This study empirically examined the influence of ownership structure(public/private) and revenue structure (high dependency on advertisement/low dependency on advertisement) of broadcasting companies upon journalists' perceptions regarding the degree of freedom in news reporting, the factors that limit freedom of press, satisfaction with the work autunomy and satisfaction with company's policies of news reporting and editing in light of resource dependency theory. The result showed that both the ownership structure and revenue structure of broadcasting companies influenced the journalists' perception regarding the degree of freedom in news gathering and reporting. It was also confirmed that only the revenue structure affected the level of satisfaction with work autonomy while the degree of satisfaction with the company's policies on editing and programing was influenced by the ownership structure. The perception of the factors that limit the freedom of press was found to be influenced only by the ownership structure, as did the satisfaction with policies on editing and programing. These results are very significant as they show that the level of dependence on advertising revenue affects journalists' news reporting, and that journalists of public broadcasting companies are influenced by the dependence on government or other political powers as well as on advertising revenue.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.3
• /
• pp.62-67
• /
• 1999

In this paper, we investigates a reliable model of the Predictive Recurrent Neural Network for the speech recognition. Predictive Neural Networks are modeled by syllable units. For the given input syllable, then a model which gives the minimum prediction error is taken as the recognition result. The Predictive Neural Network which has the structure of recurrent network was composed to give the dynamic feature of the speech pattern into the network. We have compared with the recognition ability of the Recurrent Network proposed by Elman and Jordan. ETRI's SAMDORI has been used for the speech DB. In order to find a reliable model of neural networks, the changes of two recognition rates were compared one another in conditions of: (1) changing prediction order and the number of hidden units: and (2) accumulating previous values with self-loop coefficient in its context. The result shows that the optimum prediction order, the number of hidden units, and self-loop coefficient have differently responded according to the structure of neural network used. However, in general, the Jordan's recurrent network shows relatively higher recognition rate than Elman's. The effects of recognition rate on the self-loop coefficient were variable according to the structures of neural network and their values.