• Title/Summary/Keyword: Learned Society

Search Result 1,540, Processing Time 0.03 seconds

A Method of Eye and Lip Region Detection using Faster R-CNN in Face Image (초고속 R-CNN을 이용한 얼굴영상에서 눈 및 입술영역 검출방법)

  • Lee, Jeong-Hwan
    • Journal of the Korea Convergence Society
    • /
    • v.9 no.8
    • /
    • pp.1-8
    • /
    • 2018
  • In the field of biometric security such as face and iris recognition, it is essential to extract facial features such as eyes and lips. In this paper, we have studied a method of detecting eye and lip region in face image using faster R-CNN. The faster R-CNN is an object detection method using deep running and is well known to have superior performance compared to the conventional feature-based method. In this paper, feature maps are extracted by applying convolution, linear rectification process, and max pooling process to facial images in order. The RPN(region proposal network) is learned using the feature map to detect the region proposal. Then, eye and lip detector are learned by using the region proposal and feature map. In order to examine the performance of the proposed method, we experimented with 800 face images of Korean men and women. We used 480 images for the learning phase and 320 images for the test one. Computer simulation showed that the average precision of eye and lip region detection for 50 epoch cases is 97.7% and 91.0%, respectively.

Application of reinforcement learning to hyper-redundant system Acquisition of locomotion pattern of snake like robot

  • Ito, K.;Matsuno, F.
    • Proceedings of the Korea Inteligent Information System Society Conference
    • /
    • 2001.01a
    • /
    • pp.65-70
    • /
    • 2001
  • We consider a hyper-redundant system that consists of many uniform units. The hyper-redundant system has many degrees of freedom and it can accomplish various tasks. Applysing the reinforcement learning to the hyper-redundant system is very attractive because it is possible to acquire various behaviors for various tasks automatically. In this paper we present a new reinforcement learning algorithm "Q-learning with propagation of motion". The algorithm is designed for the multi-agent systems that have strong connections. The proposed algorithm needs only one small Q-table even for a large scale system. So using the proposed algorithm, it is possible for the hyper-redundant system to learn the effective behavior. In this algorithm, only one leader agent learns the own behavior using its local information and the motion of the leader is propagated to another agents with time delay. The reward of the leader agent is given by using the whole system information. And the effective behavior of the leader is learned and the effective behavior of the system is acquired. We apply the proposed algorithm to a snake-like hyper-redundant robot. The necessary condition of the system to be Markov decision process is discussed. And the computer simulation of learning the locomotion is demonstrated. From the simulation results we find that the task of the locomotion of the robot to the desired point is learned and the winding motion is acquired. We can conclude that our proposed system and our analysis of the condition, that the system is Markov decision process, is valid.

  • PDF

A study on the status of menstruation of elementary school girls (국민학교(國民學校) 아동들의 생리실태(生理實態)에 관(關)한 조사(調査)연구)

  • Ko, Young Aie
    • Journal of the Korean Society of School Health
    • /
    • v.2 no.2
    • /
    • pp.28-35
    • /
    • 1989
  • The purpose of this study is to identify the status of menstruation of elementary schoolgirls. The date was collected from 20th May to 3rd June, 1989. The survey was conducted by using prepared questionair and 859 6th grade schoolgirls in 7 public elementary schools in Seoul were sampled for this study. The results of study are summerized as follows. 1. 13.9 percent of the school girls have experienced menarche. 0.7 percent of them began in 4th grade, 5.8 percent began in 5th grade and 7.0 percent of them began In 6th grade. 2. Among them 90.8 percent had knowledge of menstruation before menarche.43.7 percent of the respondents learned it through their mothers and 27.8 percent learned it through teachers. 3. About the menstruation, the girls who think it as a physiological phenomena as a woman showed 52.9 percent and 14.3 percent of them thought it was trouble some but, being a woman, regarded it as a inevitable thing. The repondents who think menstruation is tiresome and want not to have it were 31.9 percent. 4. At menarche 20.2 percent of the girls managed it by themselves, 72 percent disposed of it by asking someone and 2.5 percent of them managed it by their family members when the sign was detected. 5. During the menstruation 72.3 percent of the samples had some symptom and the most common symptom was abdominal pain(84.9 percent). 6. In menarche only 16.8 percent of the girls had positive response. 7. In their present menstruation 35.3 percent of them had positive response on it.

  • PDF

Factors Related to Dental Caries in Deciduous Teeth of Preschool Children (미취학아동의 유치우식 관련 요인)

  • Kim, Young-Sun;Cho, Myung-Sook;Kim, Ji-Hwa
    • Journal of Korean society of Dental Hygiene
    • /
    • v.1 no.2
    • /
    • pp.213-225
    • /
    • 2001
  • Objectives: This study was taken to investigate factors related to dental caries in deciduous teeth of preschool children. Methods : this study was completed by oral examination on 623 kindergarten pupils and questionnaire on their mothers in Sangju city from April 1 to May 20, 2001. Study design is Cross-sectional Study. Results: The mean values for decayed teeth indexed(dt index), filled teeth index(ft index), and decayed and filled teeth index(dft index) was 2.46, 1.85, and 4.30 respectively. 2.05 for girl was significantly lower than 2.77 of boy in gender (p<0.01). As child ren got older, values were higher(0.7 and 2.84 of three years old, 1.6 and 4.01 of four, 2.42 and 5.02 of five) in ft and dft index(p<0.001). Also, Mother age was almost same, that is, 4.12 dft value for 30~34 years was smaller than 6.17 for above forties(p<0.001). 2.07 and 4.68 for mothers who graduated from high school were higher than 1.36 and 3.34 for university in the ft and dft index(p<0.05), that is, the more they had a high educational level, the lower they got a values. Value 2.76 for children who had non cariogenic food was lower than 5.11 for those who had cariogenic food in dft index(p<0.001). In dft index, 3.82 of children who have learned the education on oral health from mother frequently have lower than 45 of children who have never learned (p<0.05). The ft value of children who had a snack food with mother(1.29) was a smaller than children had it at out of home(1.97)(p<0.05). The variables related 10 dft index were children and mothers age, educational level, and kinds of snack food (p<0.05). Conclusion : This study reveals that younger girl, younger mother, higher educational level, and non cariogenic foods showed lower dft index.

  • PDF

Characteristics of Highschool Students about Choosing a University (고교생들의 대학진학선택에 대한 특성)

  • Roh, Hyo-Lyun
    • Journal of the Korea Academia-Industrial cooperation Society
    • /
    • v.13 no.7
    • /
    • pp.2964-2970
    • /
    • 2012
  • The objective of this paper is to understand the characteristics of highschool students in selecting a career. The subject included 750 students from, who became a university student in 2011. The most of number of students answered that they applied to 3 to 4 universities and they decided their major and university on their last year of high school. Many placed weight on the major rather than the school. Among types of selecting career, family-type received the most voted. In selecting a major, the most answered that they consider their aptitude and then career opportunity. Most students said they were equipped with information and knowledge about the major they applied for beforehand. In selecting a university, many mainly applied for the ones with the highest chance of getting accepted. Many said they learned of the university from their high school/teacher or through the internet. A very few answered that they learned of the university through media such as TV/radio and newspaper. Therefore, appropriate information should be provided to the ones who have the greatest influence on students in selecting a career, family. Also, universities need to develop a promotion strategy that utilizes high schools as well as the internet and smartsphones.

A Study on the Publication Model of Electronic Journals for Scholarly Journals Issued at Learned Societies in Korea - The Case of Article Contribution and Electronic Journal Generation System at KISTI - (국내 학회발행 학술잡지의 전자저널 출판모형 개발에 관한 연구 - KISTI의 논문투고 및 전자저널 출판시스템을 중심으로 -)

  • Choi, Jae-Hwang
    • Journal of the Korean Society for Library and Information Science
    • /
    • v.36 no.2
    • /
    • pp.39-54
    • /
    • 2002
  • In general, researchers submit their research products to scholarly journals in order to be recognized by their colleagues, and, as a result, those are accumulated in their belonged teamed societies. However, it is a fact that the majority of teamed societies have been ignored digitized information new even though they could acquire original word files. Information flow process needs to be systematized based on digital for the research products. The purpose of this study u to develop electronic Journal generation system for scholarly journals issued at learned societies in Korea. From this study, It would be possible to deliver researcher' work directly to end users via internet, without acquisition of printed scholarly journals. For this study, KISTI's article contribution and electronic journal generation system was discussed. It also compared with other major electronic journal publication models in foreign countries.

3 Steps LVQ Learning Algorithm using Forward C.P. Net. (Forward C-P. Net.을 이용한 3단 LVQ 학습알고리즘)

  • Lee Yong-gu;Choi Woo-seung
    • Journal of the Korea Society of Computer and Information
    • /
    • v.9 no.4 s.32
    • /
    • pp.33-39
    • /
    • 2004
  • In this paper. we design the learning algorithm of LVQ which is used Forward Counter Propagation Networks to improve classification performance of LVQ networks. The weights of Forward Counter Propagation Networks which is between input layer and cluster layer can be learned to determine initial reference vectors by using SOM algorithm and to learn reference vectors by using LVQ algorithm. Finally. pattern vectors is classified into subclasses by neurons which is being in the cluster layer, and the weights of Forward Counter Propagation Networks which is between cluster layer and output layer is learned to classify the classified subclass, which is enclosed a class. Also. kr the number of classes is determined, the number of neurons which is being in the input layer, cluster layer and output layer can be determined. To prove the performance of the proposed learning algorithm. the simulation is performed by using training vectors and test vectors that ate Fisher's Iris data, and classification performance of the proposed learning method is compared with ones of the conventional LVQ, and it was a confirmation that the proposed learning method is more successful classification than the conventional classification.

  • PDF

The Hybrid LVQ Learning Algorithm for EMG Pattern Recognition (근전도 패턴인식을 위한 혼합형 LVQ 학습 알고리즘)

  • Lee Yong-gu;Choi Woo-Seung
    • Journal of the Korea Society of Computer and Information
    • /
    • v.10 no.2 s.34
    • /
    • pp.113-121
    • /
    • 2005
  • In this paper, we design the hybrid learning algorithm of LVQ which is to perform EMG pattern recognition. The proposed hybrid LVQ learning algorithm is the modified Counter Propagation Networks(C.p Net. ) which is use SOM to learn initial reference vectors and out-star learning algorithm to determine the class of the output neurons of LVa. The weights of the proposed C.p. Net. which is between input layer and subclass layer can be learned to determine initial reference vectors by using SOM algorithm and to learn reference vectors by using LVd algorithm, and pattern vectors is classified into subclasses by neurons which is being in the subclass layer, and the weights which is between subclass layer and class layer of C.p. Net. is learned to classify the classified subclass. which is enclosed a class . To classify the pattern vectors of EMG. the proposed algorithm is simulated with ones of the conventional LVQ, and it was a confirmation that the proposed learning method is more successful classification than the conventional LVQ.

  • PDF

Minority Protect Policies and Social Enterprises : Focusing on North Korean Defectors (소수자 보호정책과 사회적기업 : 북한이탈주민을 중심으로)

  • Jung, Seok-Hwan
    • The Journal of the Korea Contents Association
    • /
    • v.15 no.12
    • /
    • pp.204-211
    • /
    • 2015
  • This study made fundamental approaches on the significance of social enterprises on minority protection policies. In particular, it offered discussions focused on the relationship between social enterprises and North Korean defectors which greatly affect minority policies. Study results can be summed up as follows. First, the greatest difficulties for North Korean defectors in their adjustment to Korean society are economic activities and a sense of social differences and social enterprises are a policy which can resolve these problems in the most effective way. Second, existing social enterprises were excessively run by the central government and they were support models which lacked creativity and development. To resolve such problems, a lesson was learned that these must change into policy models with local coproduction. Third, a lesson was learned that social enterprises offer more than simple economic activity, offering the proper functions of society, culture, living, and language, and thus the participation of North Korean defectors in social enterprises should be made mandatory in certain sectors. Such research results will contribute to theoretically explaining the relationship between North Korean defectors and social enterprises as minority policies.

Preliminary PINC(Program for the Inspection of Nickel Alloy Components) RRT(Round Robin Test) - Pressurizer Dissimilar Metal Weld -

  • Kim, Kyung-Cho;Kang, Sung-Sik;Shin, Ho-Sang;Chung, Ku-Kab;Song, Myung-Ho;Chung, Hae-Dong
    • Journal of the Korean Society for Nondestructive Testing
    • /
    • v.29 no.3
    • /
    • pp.248-255
    • /
    • 2009
  • After several damages by PWSCC were found in the world, USNRC and PNNL(Pacific Northwest National Laboratory) started the research on PWSCC under the project name of PINC. The aim of the project was 1) to fabricate representative NDE mock-ups with flaws to simulate PWSCCs, 2) to identify and quantitatively assess NDE methods for accurately detecting, sizing and characterizing PWSCCs, 3) to document the range of locations and morphologies of PWSCCs and 4) to incorporate results with other results of ongoing PWSCC research programs, as appropriate. Korea nuclear industries have also been participating in the project. Thermally and mechanically cracked-four mockups were prepared and phased array and manual ultrasonic testing(UT) techniques were applied. The results and lessons learned from the preliminary RRT are summarized as follows: 1) Korea RRT teams performed the RRT successfully. 2) Crack detection probability of the participating organizations was an average 87%, 80% and 80% respectively. 3) RMS error of the crack sizing showed comparatively good results. 4) The lessons learned may be helpful to perform the PINC RRT and PSI /ISI in Korea in the future.