밀림관 형상에 따른 배관내 열성층 현상 연구

조인수(In-Su Jo),양준석(Jun-Seog Yang),박치용(Chi-Yong Park) 대한기계학회 2010 대한기계학회 춘추학술대회 Vol.2010 No.5

지적장애 특수학교장의 리더십 특성에 따른 특수교사의 직무만족도

조인수 ( In Su Jo ),김영학 ( Yeong Hak Kim ) 한국특수아동학회 2013 특수아동교육연구 Vol.15 No.1

스킵연결이 적용된 오토인코더 모델의 클러스터링 성능 분석

조인수 ( In-su Jo ),강윤희 ( Yunhee Kang ),최동빈 ( Dong-bin Choi ),박용범 ( Young B. Park ) 한국정보처리학회 2020 정보처리학회논문지. 소프트웨어 및 데이터 공학 Vol.9 No.12

오토인코더의 데이터 복원(Output result) 기능을 이용한 노이즈 제거 및 초해상도와 같은 연구가 진행되는 가운데 오토인코더의 차원 축소 기능을 이용한 클러스터링의 성능 향상에 대한 연구도 활발히 진행되고 있다. 오토인코더를 이용한 클러스터링 기능과 데이터 복원 기능은 모두 동일한 학습을 통해 성능을 향상시킨다는 공통점이 있다. 본 논문은 이런 특징을 토대로, 데이터 복원 성능이 뛰어나도록 설계된 오토인코더 모델이 클러스터링 성능 또한 뛰어난지 알아보기 위한 실험을 진행했다. 데이터 복원 성능이 뛰어난 오토인코더를 설계하기 위해서 스킵연결(Skip connection) 기법을 사용했다. 스킵연결 기법은 기울기 소실(Vanishing gradient)현상을 해소해주고 모델의 학습 효율을 높인다는 장점을 가지고 있을 뿐만 아니라, 데이터 복원 시 손실된 정보를 보완해 줌으로써 데이터 복원 성능을 높이는 효과도 가지고 있다. 스킵연결이 적용된 오토인코더 모델과 적용되지 않은 모델의 데이터 복원 성능과 클러스터링 성능을 그래프와 시각적 추출물을 통해 결과를 비교해 보니, 데이터 복원 성능은 올랐지만 클러스터링 성능은 떨어지는 결과를 확인했다. 이 결과는 오토인코더와 같은 신경망 모델이 출력된 결과 성능이 좋다고 해서 각 레이어들이 데이터의 특징을 모두 잘 학습했다고 확신할 수 없음을 알려준다. 마지막으로 클러스터링의 성능을 좌우하는 잠재변수(latent code)와 스킵연결의 관계를 분석하여 실험 결과의 원인에 대해 파악하였고, 파악한 결과를 통해 잠재변수와 스킵연결의 특징정보를 이용해 클러스터링의 성능저하 현상을 보완할 수 있다는 사실을 보였다. 이 연구는 한자 유니코드 문제를 클러스터링 기법을 이용해 해결하고자 클러스터링 성능 향상을 위한 선행연구이다. In addition to the research on noise removal and super-resolution using the data restoration (Output result) function of Autoencoder, research on the performance improvement of clustering using the dimension reduction function of autoencoder are actively being conducted. The clustering function and data restoration function using Autoencoder have common points that both improve performance through the same learning. Based on these characteristics, this study conducted an experiment to see if the autoencoder model designed to have excellent data recovery performance is superior in clustering performance. Skip connection technique was used to design autoencoder with excellent data recovery performance. The output result performance and clustering performance of both autoencoder model with Skip connection and model without Skip connection were shown as graph and visual extract. The output result performance was increased, but the clustering performance was decreased. This result indicates that the neural network models such as autoencoders are not sure that each layer has learned the characteristics of the data well if the output result is good. Lastly, the performance degradation of clustering was compensated by using both latent code and skip connection. This study is a prior study to solve the Hanja Unicode problem by clustering.

솔레노이드 및 피에조 인젝터의 고속분사 응답성 해석

조인수(In Su Jo),이중협(Jung Hyup Lee),이진욱(Jin Wook Lee) 대한기계학회 2012 大韓機械學會論文集B Vol.36 No.10

고압 인젝터의 성능은 디젤 연소엔진의 동력, 배출물, 연료소모와 직접적인 관계가 있다. 본 논문에서는 솔레노이드 코일과 피에조 세라믹으로 구동되는 커먼레일 디젤 분사용 인젝터의 응답 특성을 AMESim 코드를 사용하여 상대 비교 연구를 수행하였다. 따라서 연료압력, 분공경을 주요 해석변수로 설정하였다. 본 연구를 수행한 결과, 솔레노이드 구동 인젝터에 비해 피에조 구동 인젝터가 상대적으로 더 빠른 응답성과 더 높은 제어성을 가짐을 알 수 있었으며, 특히 다단분사 적용시, 이런 결과가 매우 효과적임을 확인할 수 있었다. It is well known that the performance of a diesel injector is directly related to the power, emission, and fuel consumption of the diesel combustion engine. In this study, the injection response characteristics of CRDi injectors driven by a solenoid coil and a piezoceramic were investigated by using the AMESim simulation code. Some analytical parameters such as the fuel pressure and hole diameter were considered. From this study, it was shown that the piezo-driven injector had a faster response and had better control capability than the solenoid-driven injector. In addition, it was found that the piezo-driven injector can be utilized more effectively in a multiple injection scheme than a solenoid-driven injector.

AMESim기반 CRDi용 인젝터의 2단분사 동적거동 특성해석

조인수 ( In Su Jo ),권지원 ( Ji Won Kwon ),이진욱 ( Jin Wook Lee ) 한국분무공학회 2012 한국액체미립화학회지 Vol.17 No.2

For reduction of CO, NOx and soot emission emitted by diesel diffusion combustion, the authors focused on injection actuator to improve fuel availability inside combustion chamber. In this study, it was investigated the internal dynamic characteristics of two-stage injection with diesel injectors with different driving type for the common rail direct injection by using the AMESim simulation code. The analysis parameter defined such as fuel pressure, injection hole`s diameter and driven voltage. As the results, it was shown that the piezo-driven injector had a faster response and had better control capability than the solenoid-driven injector. It was found the piezo-driven injector can be utilized effectively as multiple injector than solenoid-driven injector.

Mask R-CNN 과 zi2zi 모델을 활용하여 탐지된 객체의 스타일을 변환시키는 신경망 모델

조인수 ( In-su Jo ),최동빈 ( Dong-bin Choi ),박용범 ( Young B. Park ) 한국정보처리학회 2021 한국정보처리학회 학술대회논문집 Vol.28 No.2

스타일 변환 모델은 이미지 전체나 이미지 내에서 사용자가 지정한 영역을 대상으로 스타일을 변환시킨다. 이런 방식은 이미지 내의 다수의 객체에 대해 스타일 변환을 시행할 때 일일이 영역을 지정해 줘야 한다는 불편함과 결과물의 전체 해상도가 떨어진다는 한계를 가지고 있다. 본 논문에서는 이런 한계들을 극복하기 위해 객체탐지 모델과 스타일변환 모델을 연동한 객체스타일변환모델을 제안하고 모델 간 연동방법에 대해 자세히 서술한다. 객체탐지모델인 Mask R-CNN 을 통해 필요한 객체를 탐지하고 탐지한 객체의 특징맵들을 스타일변환 모델인 zi2zi 의 입력 값으로 전달하여 이미지 내의 필요한 객체들만 스타일변환이 이루어지도록 모델이 동작한다. 이러한 모델은 기존에 있는 두 모델을 재사용함으로써 모델을 처음부터 새로 설계할 필요가 없다는 장점이 있으며, 공개된 다양한 모델들을 서로 융합하여 사용할 수 있는 방법을 제시하는데 도움을 줄 것이다.

PEMFC 발열에 의한 Stack내 열응력 분석

조인수(In Su Jo),권오정(Oh-Jung Kwon),하정석(Jeong Seok Ha),황성하(Sung Ha Hwang),오병수(Byeong Soo Oh) 한국생산제조학회 2017 한국생산제조시스템학회 학술발표대회 논문집 Vol.2017 No.12

funcGNN과 Siamese Network의 코드 유사성 분석 성능비교

최동빈,조인수,박용범,Choi, Dong-Bin,Jo, In-su,Park, Young B. 한국반도체디스플레이기술학회 2021 반도체디스플레이기술학회지 Vol.20 No.3

As artificial intelligence technologies, including deep learning, develop, these technologies are being introduced to code similarity analysis. In the traditional analysis method of calculating the graph edit distance (GED) after converting the source code into a control flow graph (CFG), there are studies that calculate the GED through a trained graph neural network (GNN) with the converted CFG, Methods for analyzing code similarity through CNN by imaging CFG are also being studied. In this paper, to determine which approach will be effective and efficient in researching code similarity analysis methods using artificial intelligence in the future, code similarity is measured through funcGNN, which measures code similarity using GNN, and Siamese Network, which is an image similarity analysis model. The accuracy was compared and analyzed. As a result of the analysis, the error rate (0.0458) of the Siamese network was bigger than that of the funcGNN (0.0362).

포스터 전시회 : 서보 유압형 피에조 인젝터 모델 개발 및 2단 분사 검증

김상명 ( Sang Myeong Kim ),정명철 ( Myoung Chul Jeong ),성기수 ( Gi Su Sung ),조인수 ( In Su Jo ),이진욱 ( Jin Wook Lee ) 한국액체미립화학회 2013 한국액체미립화학회 학술강연회 논문집 Vol.2013 No.-

Performance of diesel engine with high fuel pressure injection is related its emission gas and fuel consumption. In this study, by using 3rd generation piezo-driven injector model, two-stage injection characteristics and effects were analyzed. And numerical model with injection rate and hydraulic pressure about two-stage injection were developed and verified with experimental results.

精神薄弱兒의 敎育要求分析

趙寅洙 又石大學校 1982 論文集 Vol.4 No.-

This study aims at the search for educational requirement of the Mentally Retarded, especially focusing on E M R and T M R. It also attempts to ask for educational requirement through researching the materials which have been studied, home and abroad, about the Mentally Retarded -- their educational possibility; their developmental characteristics; and other characteristics of the physical movement, social emotion and perceptive learning. The educational requirement, developmental possibility and other particular characteristics of the Mentally Retarded are briefly as follows: 1. Developmental Possibility of the Mentally Retarded. 1) In general, development is possible on the level of mental age 8-12 for the E M R, 4-8 for the TMR 2. Characteristics of the Mentally Retarded. 1) Perceptive Characteristic: For the Mentally Retarded, it is hard to generalize learning. There is much unbalance in the low grade factors of intellectuality -- especially, speech factor shows more inferior than motion factor. When mental ages are the same, the perceptions of the Mentally Retarded are more superior to the concrete object, but more inferior to the abstract object than those of the normal. It is also a characteristic to develope speech ability slowly. 2) Socio-Emotional Characteristic They are exceedingly retarded in the field of speech in their social lives, but less retarded comparatively in the concrete works. They are the same with the normal from character and emotion, but are not able to develop the field of ego and super-ego. It is a characteristic that the solidity of emotional function fields is high in the structural phase. 3) Physical, Motional Characteristic They are not much retarded in the physical growth, but are much more retarded for the sitting or standing alone, taking off one step or walking from compared with the physical sizes. In the motion ability, they are almost below-10 in the all fields -- that is; balance function, whole body exercise ability, fingers-correspondence, imitation of seperation, running, broad jumping, soft-ball drawing etc. 4) Characteristic of Learning. The Learning abilities of EMR are generally low. So, the degree of elementary school 6 grade EMR is that of kidergarten at the end of learning age, and they will be able to have the ability of about 3 grade normal student. For TMR, it is possible to train the ability of communication, but is almost impossible to teach general letter learning (art function). In conclusion sellection and organization of all educational subject matters must be done according to the educational requirement and characteristics of the Mentally Retarded. And the followings have to be emphasized 1. The contents, which enrich them with good citizen characters, must be chosen, considering of the development possibility of learning ability. 2. The contents have to be formed with those which can be used in present life directly, can be transformed well and can improve the quality of life. 3. The contents, which the life needs, and become the foundation of future vocational life, and increase the aptitude, must be established. 4. The contents of all the education have to be organized with small steps of guidance which is interesting and stressed on the repetition quality.