• 한국멀티미디어학회논문지
• /
• 제22권1호
• /
• pp.1-9
• /
• 2019

Single image rain removal is a typical inverse problem which decomposes the image into a background scene and a rain streak. Recent works have witnessed a substantial progress on the task due to the development of convolutional neural network (CNN). However, existing CNN-based approaches train the network with synthetically generated training examples. These data tend to make the network bias to the synthetic scenes. In this paper, we present an unsupervised framework for removing rain streaks from real-world rainy images. We focus on the natural phenomena that static rainy scenes capture a common background but different rain streak. From this observation, we train siamese network with the real rain image pairs, which outputs identical backgrounds from the pairs. To train our network, a real rainy dataset is constructed via web-crawling. We show that our unsupervised framework outperforms the recent CNN-based approaches, which are trained by supervised manner. Experimental results demonstrate that the effectiveness of our framework on both synthetic and real-world datasets, showing improved performance over previous approaches.

• 전자공학회논문지B
• /
• 제33B권8호
• /
• pp.96-104
• /
• 1996

In this paper, we propose a new algorithm that obtains the surfac eorientation of the texture image using structural approaches. The proposed method showed that structural approaches can be effectively used in 3-D analysis of textures as well as description and segmentation without additional information. By examining fourier power spectrum of the texture image, we detemine the tilt of the textured surface. Then, 1-D projection information of the texture in the obtained tilt direction is used to compute the slant. Using the obtained information, we can compute the vanishing point, and rearrange the textured surface with lines converging to the vanishing point and lines perpendicular to the tilt direction. In the experimental results, we have ascertained the proposed algorithm can make a rpecise 3-D analysis of structural textures.

• 한국해양공학회지
• /
• 제31권2호
• /
• pp.91-102
• /
• 2017

Recently, a new type of LNG membrane Tank called the "KC-1 membrane LNG Tank" was developed by KOGAS (Korean Gas Corporation). It is necessary to estimate the temperature distribution of the hull structure and insulation system for this new LNG tank, as well as the BOR (Boil-Off Rate) when exposed to outside temperature conditions to ensure the integrity of the tank structure and limit LNG evaporation, from a safety evaluation point of view. In this study, temperature distribution calculations for the hull structure and insulation system of the KC1 membrane tank were compared by employing four numerical approaches under the IGC condition. Approaches 1-3 studied 2D simulations and approach 4 used a 3D numerical simulation. Approach 1 was calculated by in-house Excel VBA codes and the three other approaches utilized ANSYS Fluent. The BOR of approach 4, the 3D simulation case, for the IGC condition was 0.0986%/day.

• 통합자연과학논문집
• /
• 제5권1호
• /
• pp.1-5
• /
• 2012

Quantitative structure-activity relationship (QSAR) methodologies have been applied for many years, to correlate the relationship between physicochemical properties of chemical substances and their biological activities to generate a statistical model for prediction of the activities of new chemical entities. The basic principle behind the QSAR models is that, how structural variation is responsible for the difference in biological activities of the compounds. 3D-QSAR has emerged as a natural extension to the classical Hansch and Free-Wilson approaches, which develops the 3D properties of the ligands to predict their biological activities using various chemometric techniques (PLS, G/PLS, ANN etc). It has served as a valuable predictive tool in the design of pharmaceuticals and agrochemicals. This review seeks to provide different 3D-QSAR approaches involved in drug designing process to develop structure-activity relationships and also discussed the fundamental limitations, as well as those that might be overcome with the improved methodologies.

• 기술혁신학회지
• /
• 제8권1호
• /
• pp.220-236
• /
• 2005

In today's R&D environment, as the complexity of IT systems increases and the expected quality level for R&D outcomes enlarges, it has been critical issues for us to effectively define and improve the R&D standard processes tailored for our organization and projects. To meet this needs, we defined the R&D standard processes based on the international standards for systems and software engineering and our existing traditional work methods, and then applied them of our ongoing projects. Now, we are improving our defined R&D standard processes continuously by monitoring of the processes and products and through the feedback of the application results. In this paper, we introduce the R&D standard process model developed by our organization (E research Institute in Korea) and the requirements and approaches used for their development and improvement. We compare our defined R&D standard processes with ISO/IEC 15288 which is the international standards for systems engineering and summarize the lessons learned through their development and improvement.

• Journal of Information Display
• /
• 제8권1호
• /
• pp.26-31
• /
• 2007

This paper presents ergonomic approaches to editing and evaluating content for 3D displays. Two systems, a nonlinear editing system and an evaluation system, were developed to improve viewing safety and comfort for 3D content observers and provide better usability for the creators from the viewpoint of ergonomics.

• International Journal of Industrial Entomology and Biomaterials
• /
• 제1권2호
• /
• pp.79-86
• /
• 2000

Strengthening of R & D components of sericulture in India since the sixties has led to a quantum jump in silk production and presently India is the second largest producer in the world. This achievement is primarily due to a number of breakthroughs in R & D to match the requirements of tropics, by way of introduction of improved mulberry varieties and silkworm breeds, better mulberry cultivation and rearing management practices suited to tropical conditions. Of late, new approaches in molecular biology and biotechnology have also been vigorously pursued to strengthen the current conventional strategies. The present paper attempts to provide an overview of the present status of silk production in both mulberry and non-mulberry sectors, breakthroughs achieved through new approaches of biotechnology and the future prospects for maximizing silk productivity in India.

• 한국통신학회논문지
• /
• 제36권1B호
• /
• pp.101-110
• /
• 2011

광대역 네트워크 기술의 발전함에 따라 점점 더 많은 클라이언트들이 여러 가지 VoD (Video on Demand) 서비스를 사용할 수 있게 되었다. 많은 클라이언트들을 서포트하기 위하여, VoD 전송 방식을 설계할 때 아래의 몇가지 요소들을 고려하여야 한다. 즉 사용자 대기 시간 (Viewer's Waiting Time), 매 클라이언트에서 요구하는 최대 버퍼 요구량 (Buffer Requirement at Each Client ), 비디오 전송에 요구되는 채널개수(number of channel required for video delivery)와 비디오 분할 복잡도(video segmentation complexity) 등이다. 최근의 여러 가지 VoD 서비스들 중에서 Polyharmonic과 Staircase전송 방식이 사용자 대기 시간과 매 클라이언트에서 요구하는 최대 버퍼 요구량에서 제일 좋은 성능을 보여주고 있다. 그러나 이런 방식들은 하나의 비디오를 너무 많은 세그먼트들로 나누어야 하는데 이런 방식은 동시에 관리하고 사용하는 태널의 개수가 많아지게 한다. 이런 문제들을 해결하기 위하여 이 논문에서는 시스템 복잡도를 낮추면서 사용자 대기시간과 최대 버퍼 요구량의 성능을 동시에 향상시키기 위한 방법으로 Polyharmonic과 Staircase 모델을 헤드 파트에 적용하고 기존의 VoD 전송 방식들 가운데 가장 간단한 모델인 Staggered을 뒤 파트에 적용하여 Polyharmonic-Staircase-Staggered (PSS) 방식을 제안하고 있다. 이 방식은 간단하고 효율적이다. 수학적분석을 통해 사용자 대기시간은 기존의 Harmonic Broadcasting 방식과 비교했을 때 대역폭을 조금만 더 크게 차지하면 거의 동일한 성능을 나타내고 있고 최대 버퍼 요구량은 비디오 분할 계수의 조절에 의해 Harmonic Broadcasting보다 60% 우수한 성능을 나타냄을 알 수 있다. 제일 중요한 것은 제안된 방식은 제안된 방식은 실제 용용에서 중요한 요소인 비디오 서브세그먼트의 수, 동시에 관리하는 채널의 수, 동시에 사용하는 채널의 수도 크게 감소시켰다는 것이다. 또한 다양한 환경에 따라 비디오 분할 계수를 어떻게 적절히 조정하는지도 서술하고 있다.

• International Journal of Computer Science & Network Security
• /
• 제23권9호
• /
• pp.186-191
• /
• 2023

Neural Networks are widely used for huge variety of tasks solution. Machine Learning methods are used also for signal and time series analysis, including electrocardiograms. Contemporary wearable devices, both medical and non-medical type like smart watch, allow to gather the data in real time uninterruptedly. This allows us to transfer these data for analysis or make an analysis on the device, and thus provide preliminary diagnosis, or at least fix some serious deviations. Different methods are being used for this kind of analysis, ranging from medical-oriented using distinctive features of the signal to machine learning and deep learning approaches. Here we will demonstrate a neural network-based approach to this task by building an ensemble of 1D CNN classifiers and a final classifier of selection using logistic regression, random forest or support vector machine, and make the conclusions of the comparison with other approaches.

• 대한조선학회논문집
• /
• 제53권2호
• /
• pp.127-134
• /
• 2016

This paper presents practical issues for test and evaluation(T&E) methods to develop defense systems. Our argument is motivated by several domestic defense cases and the cases lead us to discuss two main factors for reliable defense systems development: 1) statistical approaches and 2) technical schemes. Specifically, statistical approaches enable to provide credible interpretations about T&E results in the decision-making process. With practical T&E results of the “Red Shark” torpedo, we performed statistical hypothesis tests and suggest a minimum sample size to accept the hypothesis. Next, technical schemes have more direct effects on improving reliability of developed defense systems and we shortly introduce tools development for systems verification that is required to integrate several sub-systems, e.g., combat, sensor, weapon, and communication systems, within a defense system. We additionally summary some domain cases using modeling and simulation techniques for successful T&E. In closing, we expect that the paper shows empirical investigation and lessons learned with these two practical issues, which provides a guide those who desire to make decisions about reliable defense systems development.