http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
치근면 처치시 결합조직 부착에 관한 초기효과의 조직병리학적 연구
문상준,이종헌,이재현 단국대학교 치의학연구소 1993 논문집 Vol.5 No.-
An essential fact in the regeneration of new periodontal tissue after periodontal therapy is the reattachment of collagen fibers to the tooth. Two phenomena play a fundamental role in preventing new connective tissue attachment to the exposed root surface ; 1) The apical migration of the junctional epithelium 2) The contamination of cementum by toxic substances, especially endotoxins. Authors have used rat submucosal implantation of root sections to study the connective tissue healing to periodontally diseaed root, previously planed and demineralized with citric acid and tetracycline-HCI. The results were obtained as follows. 1. The connective tissue attachment was increased in tetracycline, citric acid, non disease, scaling and root planing order and inflammatory reaction was seen in the rat teeth, no treatment group. 2. Collagen fiber attachment at the dentin surface was more increased than cementum surface. 3. In 2 week of citric acid and tetracycline-HCI specimens, osteoid was seen near the fibrotic band. 4. In the MT view, collagen fiber formation was increased with time and the numerous collagen fiber and connective tissue was more densly attached to the tooth surfaces in the tetracycline-HCI group than the citric acid group.
문상준,전종준 한국데이터정보과학회 2017 한국데이터정보과학회지 Vol.28 No.6
The online learning is a process of obtaining the solution for a given objective function where the data is accumulated in real time or in batch units. The stochastic gradient descent method is one of the most widely used for the online learning. This method is not only easy to implement, but also has good properties of the solution under the assumption that the generating model of data is homogeneous. However, the stochastic gradient method could severely mislead the online-learning when the homogeneity is actually violated. We assume that there are two heterogeneous generating models in the observation, and propose the a new stochastic gradient method that mitigate the problem of the heterogeneous models. We introduce a robust mini-batch optimization method using statistical tests and investigate the convergence radius of the solution in the proposed method. Moreover, the theoretical results are confirmed by the numerical simulations. 온라인 학습은 자료가 실시간으로 혹은 배치 단위로 축적되는 상황에서 주어진 목적함수의 해를 계산하는 방법을 말한다. 온라인 학습 알고리즘 중 배치를 이용한 확률적 경사 하강법 (stochastic gradient decent method) 은 가장 많이 사용되는 방법 중 하나다. 이 방법은 구현이 쉬울 뿐만 아니라 자료가 동질적인 분포를 따른다는 가정 하에서 그 해의 성질이 잘 연구되어 있다. 하지만 자료에 특이값이 있거나 임의의 배치가 확률적으로 이질적 성질을 가질 때, 확률적 경사 하강법이 주는 해는 큰 편이를 가질 수 있다. 본 연구에서는 이러한 비정상 배치 (abnormal batch) 있는 자료 하에서 효과적으로 온라인 학습을 수행할 수 있는 수정된 경사 하강 알고리즘 (modified gradient decent algorithm)을 제안하고, 그 알고리즘을 통해 계산된 해의 수렴성을 밝혔다. 뿐만 아니라 간단한 모의 실험을 통해 제안한 방법의 이론적 성질을 실증하였다.
경량화 U-Net을 사용한 엣지 디바이스용 미세플라스틱 이미지 분할
문상준,최다솜,이광희,정지현,이혜미,박선욱,장경선 한국정보과학회 2024 정보과학회 컴퓨팅의 실제 논문지 Vol.30 No.4
해양 생태계에 부적절한 영향을 끼치는 미세플라스틱을 검출하여 배출되는 양을 줄이는 작업이 필요하다. 미세플라스틱은 현미경 이미지 분할 딥러닝 모델을 적용해 검출할 수 있는데, 정확도를 높이기 위해서는 데이터의 다운샘플링을 지양하는 것이 중요하다. 하지만, 이를 위해서는 일반적으로 모델의 파라미터 및 연산량이 증가하는 문제가 있어 휴대형 장치에 실시간으로 동작하기 어렵다. 따라서 딥러닝 모델을 휴대형 장치를 사용해 현장에서 채취한 시료로부터 즉각 미세플라스틱을 검출하는 과정에는 적용하기 위해서는 데이터의 해상도를 유지하되 경량화 된 모델이 필요하다. 모델 경량화 실험을 위해서 이미지 분할 분야의 대표적인 모델인 U-Net의 합성곱 계층에 기존의 다양한 경량화 방법을 적용하여 U-Net과의 성능을 비교하였다. Depthwise Separable 합성곱 계층을 사용한 경우 U-Net에 비하여 유사한 mIoU 성능을 보이며, 82.6% 감소된 연산량을 보였다. 또 행렬 분해 및 Group 합성곱을 적용한 경우는 U-Net에 비해 7.41%의 mIoU 성능 하락이 있었지만, 99.55% 감소된 연산량을 보인다. 이는 현장에서 즉각적으로 데이터를 취득해야 하는 휴대형 장치에 적용해 이점을 얻을 수 있을 것으로 생각된다. There is a need to detect microplastics that adversely affect marine ecosystems and reduce their emissions. Microplastics can be detected by applying deep learning models to segment microscopic images. While it is important to avoid downsampling the data to improve accuracy, this usually requires an increase in model parameters and computation, making it difficult to operate in real-time on mobile devices. Therefore, to apply deep learning models to the process of instant microplastic detection from samplescollected in the field using mobile devices, a lightweight model is required while maintaining the resolution of the data. For the model lightweight experiment, we applied various existing lightweight methods to the convolution layer of U-Net, a representative model in the image segmentation field, and compared its performance with U-Net. The Depthwise Separable convolution layer shows mIoU performance that is comparable to that of U-Net, with 82.6% less computation. In addition, the Matrix Factorization and Group convolution layers show 99.55% less computation, although there is a 7.41% mIoU performance drop compared to U-Net. This is thought to be beneficial for mobile devices that require immediate data acquisition in the field.
치근면 처치시 결합조직 부착에 관한 초기효과의 조직병리학적 연구
문상준,이종헌,이재현,Moon, Sang-Joon,Lee, Chong-Heon,Lee, Jae-Hyun 대한치주과학회 1994 Journal of Periodontal & Implant Science Vol.24 No.1
An essential fact in the regeneration of new periodontal tissue after periodontal therapy is the reattachment of collagen fibers to the tooth. Two phenomena play a fundamental role in preventing new connective tissue attachment to the exposed root surface ; 1) The apical migration of the junctional epithelium 2) The contamination of cementum by toxic substances, especially endotoxins. Authors have used rat submucosal implantation of root sections to study the connective tissue healing to periodontally diseaed root, previously planed and demineralized with citric acid and tetracycline- HCl. The results were obtained as follows. 1. The connective tissue attachment was increased in tetracycline, citric acid, non disease, scaling and root planing order and inflammatory reaction was seen in the rat teeth, no treatment group. 2. Collagen fiber attachment at the dentin surface was more increased than cementum surface 3. In 2 week of citric acid and tetracycline-HCl specimens, osteoid was seen near the fibrotic band. 4. In the MT view, collagen fiber formation was increased with time and the numerous collagen fiber and connective tissue was more densly attached to the tooth surfaces in the tetracycline-HCl group than the citric acid group.
실시간 홀로그래픽 간섭법을 이용한 압력용기의 내부결함 측정법
문상준,강영준,백성훈,김철중,Moon, Sang-Joon,Kang, Young-June,Baik, Sung-Hoon,Kim, Cheol-Jung 대한기계학회 1996 大韓機械學會論文集A Vol.20 No.4
Conventional measurement methods using ultrasonic wave or x-ray, eddy current for non-destructive testing(NDT) in nuclear power plants and other industrial plants have been utilized as the method of contact with objects to be inspected. For this reason these methods require relatively much time and inspection area is limited by the location of probe or film. But holograpic interferometry which is a non-contact optical measurement method using a coherent light source has an advantage that quantative measurement can be performed at a time. In this paper a new method using realtime holographic interfreometry and image processing for detecting internal flaws of pressure vessels is presented.
정신질환에 대한 일반인의 사회적 거리감에 영향을 미치는 요인
문상준,이진석,박수경,이선영,김윤,김용익,신영수,Moon, Sang-Jun,Lee, Jin-Seok,Park, Sue-Kyung,Lee, Sun-Young,Kim, Yoon,Kim, Yong-Ik,Shin, Young-Soo 대한예방의학회 2008 예방의학회지 Vol.41 No.6
Objective: The purpose of this study was to investigate impact of knowledge familiarity, and prejudice about mental illness as well as demographic factors on the social distance from mentally ill people, which is a proxy measure of discrimination. Method: To assess the impact of knowledge and familiarity, prejudice about mental illness and demographic factors on the social distance from mental illness, we conducted a telephone survey in South Korea with the responders being nationally representative people who were 18 years old or over (n=1040). Independent samples T-tests, one way ANOVA and linear regression analysis were performed to analyze the results of the survey. Result: The social distance from mental illness decreased as the knowledge and familiarity increased, but the social distance was increased as prejudice was increased. Prejudice had a greater impact on social distance than familiarity and knowledge. Females showed greater social distance than did males. A higher education level had a negative effect on social distance. Conclusion: to reduce the social distance from mentally ill people, efforts to increase the familiarity about mental illness as well as efforts to educate people about mental illness are important.