• Journal of the Korean Geographical Society
• /
• v.43 no.6
• /
• pp.894-913
• /
• 2008

This study examines scale issues in the contemporary feminist migration literature. Scale appears as important, yet poorly understood concept in this field of study. The increasing attention to the feminization of migration requires not only gendered, but also scalar-sensitive approaches. Feminists criticize the conventional approach to the migration as a gender-blind approach that privileges national scale around which migration processes are organized. Claiming multiscalar and interscalar analyses, they propose investigations ranging from macro to micro processes which include globalized gendered division of labor, transnational family networks, and reproduction which takes place in and through the bodies and homes of migrant women. The migrant women, the major actors in recent transnational migration, cross various borders: the national boundaries and the public and private divides, in particular. This crossover can unsettle patriarchal gender relations which have been established based on the physical and symbolic division of nation-states and public/private spheres. Blurring these divisions accompanies social construction of various scales. The transnational family networks of migrant women, for example, show the construction of a transnational scale by migrant women as well as globalization from below. This paper points out misunderstandings of scale in the feminist migration literature and attempts to fill the gaps by introducing the meanings and implications of scales developed mostly by feminist geographers. In so doing, it promotes the interdisciplinary communication.

• Journal of the Korean GEO-environmental Society
• /
• v.13 no.9
• /
• pp.45-51
• /
• 2012

The calcium hydroxide$(Ca(OH)_2)$ which is the cement hydrate flowed into the tunnel by groundwater is reacted with microorganism in the soil, carbon dioxide$(CO_2)$ and the vehicle's exhaust gas$(SO_3)$. So its by-products are precipitated at the drainage pipe and these cause the drainage clogging. By this phenomenon, Degradation of water flow at the drainage system of the tunnel occurred and also pore water pressure is increased. Hence the acceleration of seepage and degradation of lining is occurred. The purpose of this study is to evaluate the field applicability of the Quantum Stick and Magnetic treatment in prevention of scale deposits at the Namsan ${\bigcirc}{\bigcirc}$ tunnel and the Zone ${\bigcirc}{\bigcirc}{\bigcirc}$ of subway. These technologies were installed into drainpipes with their performance monitored through occasional site visits. SEM and XRD were also performed on scale collected from these drainpipes. As a result, in case which factor technology is applied, scale creation is remarkably decreased and especially Quantum Stick treatment performing better than Magnetic treatment. Therefore, additional application of Quantum Stick or Magnetic treatment to the existing drainage is expected to decrease the drainage clogging of the drainage.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2010.05a
• /
• pp.847-850
• /
• 2010

최근 많은 수문학자들은 전지구적 기후변화로 인한 피해 예방과 저감을 위해 기후변화가 수문학적으로 어떤 영향을 미치고 있는지 알기 위해 많은 연구를 진행하고 있으며, 기후변화시나리오를 작성하고자 이산화탄소 배출농도를 가정하여 다양한 시나리오를 생성하고 있다. 본 연구에서는 효율적인 수자원 관리를 위해 저해상도의 GCM(General Circulation Models) 모형에서 생성되는 모의결과를 유역 규모의 단위로 스케일 상세화 기법(downscaling)을 적용 시켜 보고자 한다. 이를 위해 2007년 IPCC AR4와 함께 제시된 SRES A1B 시나리오를 채택하여 우리나라 기상청이 연구에 참여 제공하고 있는 EHCO-G 모델의 모의결과를 이용하여 소양강 유역에 적용하였다. 상세화 기법으로는 현재와 과거의 입력값들과 이에 대응된 출력값들을 알고 있는 경우에 미래의 새로운 입력값들에 대한 예측값들을 추출하는데 유용하며, 비선형적 비연속적인 특성이 강한 모델에 강점을 가지고 있는 인공신경망(Artificial Neural Network) 모델을 사용하고자 한다.

• Annual Conference of KIPS
• /
• 2021.11a
• /
• pp.764-767
• /
• 2021

While Generative Adversarial Networks (GANs) have seen huge success in image synthesis tasks, synthesizing high-quality images from text descriptions is a challenging problem in computer vision. This paper proposes a method named Text-to-Face Generation Using Multi-Scale Gradients for Conditional Generative Adversarial Networks (T2F-MSGGANs) that combines GANs and a natural language processing model to create human faces has features found in the input text. The proposed method addresses two problems of GANs: model collapse and training instability by investigating how gradients at multiple scales can be used to generate high-resolution images. We show that T2F-MSGGANs converge stably and generate good-quality images.

• Journal of the Korean Geotechnical Society
• /
• v.29 no.7
• /
• pp.5-15
• /
• 2013

Clogging phenomenon is one of the important problems in deteriorated tunnels, and it caused inhibition of drainage system by long-term behavior. Clogging phenomenon is mainly composed of $CaCO_3$ in the form calcite. Calcite is generally created by the reaction of $Ca(OH)_2$ with $CO_2$ emitted from vehicles. The structure of deteriorated tunnels was simulated and the setting of outflow from drainage pipe was observed in this study. The test was experienced by changing the slope of drainage system because existing drainage system was pracitced almost below $5^{\circ}$. As a result, in case of drainage system's slope is $2^{\circ}$, Quantum Stick has an effect for prohibiting scale in drainage system, but magnetic treatment was not effective. As a result, in case of drainage system's slope is $5^{\circ}$, both technologies were effective for prohibiting scale in drainage system, but Quantum Stick was especially more effective than magnetic treatment.

• Journal of KIISE
• /
• v.43 no.8
• /
• pp.910-918
• /
• 2016

A tensor is a multi-dimensional array that represents many data such as (user, user, time) in the social network system. A tensor generator is an important tool for multi-dimensional data mining research with various applications including simulation, multi-dimensional data modeling/understanding, and sampling/extrapolation. However, existing tensor generators cannot generate sparse tensors like real-world tensors that obey power law. In addition, they have limitations such as tensor sizes that can be processed and additional time required to upload generated tensor to distributed systems for further analysis. In this study, we propose TeT, a distributed tera-scale tensor generator to solve these problems. TeT generates sparse random tensor as well as sparse R-MAT and Kronecker tensor without any limitation on tensor sizes. In addition, a TeT-generated tensor is immediately ready for further tensor analysis on the same distributed system. The careful design of TeT facilitates nearly linear scalability on the number of machines.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.9 no.3
• /
• pp.42-50
• /
• 2001

In this paper, the effects of scale formation in engine water jacket upon the thermal durability of engine itself and its component parts were studied. To understand the effect of quality of water, a full load engine endurance test for 50 hours was carried out with not-treated underground water. The followings were found through the tested engine inspection after the endurance test; 1-2 mm thick scale formation in the engine water jacket, valve seat wear, piston top land scuffing, piston pin stick, and cylinder bore scuffing in siamese area. In order to understand the causes of above test results, the heat rejection rate to coolant, the metal surface temperature of combustion chamber, and the oil and exhaust gas temperatures were measured and analyzed. The scale formed in the engine water jacket played a role as thermal insulator. The scale formed in the engine reduced the heat rejection rate to coolant and it caused to increase the metal surface temperature. The reduced heat rejection rate to coolant increased the heat rejection rate to oil and exhaust gas and increased the oil and exhaust gas temperature. Also, the reasons of valve seat wear, piston top land scuffing and cylinder bore scuffing, and piston pin stick quantitatively analyzed in this paper.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2015.07a
• /
• pp.9-12
• /
• 2015

In this paper, we propose a new feature matching algorithm by modifying and combining the FAST(Features from Accelerated Segment Test) feature detector and SURF feature descriptor which is robust to the distortion of the given image. Scale space is generated to consider the variation of the scale and determine the candidate of features in the image robust to the noise. The original FAST algorithm results in many feature points along edges. To solve this problem, we apply the principal curvatures for refining it. We also use SURF descriptor to make it robust against the variations in the image by rotation. Through the experiments, it is shown that the proposed algorithm has better performance than the conventional feature matching algorithms even though it has much less computational load. Especially, it shows a strength for noisy images.

• 한국전산유체공학회:학술대회논문집
• /
• 2008.03b
• /
• pp.109-112
• /
• 2008

In this work, a mathematical heat transfer model of a walking-beam type reheating furnace that can predict the formation and growth of the scale layer, which is produced due to oxidative reaction between the furnace oxidizing atmosphere and the steel surface in the reheating furnace, has been developed. The model can also predict the heat flux distribution within the furnace and the temperature distribution in the slab and scale throughout the reheating furnace process by considering the heat exchange between the slab and its surroundings in the furnace, including radiant heat transfer among the slabs, the skids, the hot gases and the furnace wall as well as the gas convection heat transfer in the furnace. Using the model developed in this work, the effects of the scale layer on the heat transfer characteristics and temperature behavior of the slab is investigated. A comparison is also made between the predictions of the present model and the data from an in situ measurement in the furnace, and a reasonable agreement is founded.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2015.05a
• /
• pp.421-421
• /
• 2015

2007년 IPCC 4차 기후변화 평가보고서 발간 이후, 최근 제 5차 기후변화 평가보고서에 표준 온실가스 시나리오를 대표농도경로(Representative Concentration Pathway, RCP)로 새롭게 선정하여 발간되었다. 이러한 기후변화 시나리오를 적용하여 수자원변화를 예측하는 연구들이 진행되었다. 그러나 기후변화 모델을 제공하는 기관이 많고 전지구적인 스케일로 제공되고 있어 모델의 선택여부와 지역적인 특성으로 인해 발생하는 스케일 불일치 등 다양한 불확실성을 포함하고 있다. 또한, 본 연구 대상지역인 도암댐 유역은 상류에 고랭지 밭이 다수 위치해 있으며 2018년 동계올림픽을 유치하는 평창에 속해있어 2011년 이후로 급격한 개발이 이루어졌고 지속적으로 토지 이용변화가 일어나는 유역이다. 따라서 본 연구에서는 RCP 4.5와 8.5시나리오를 대상으로 총 20개의 기후변화 모델 자료를 수집하였고 지역오차보정을 통해 지역적인 스케일의 불일치를 개선하고 미래시나리오에 대해서는 비정상성을 고려한 비정상성 분위사상법을 통해 미래 시나리오의 정확성을 높였다. 과거 토지이용변화추세를 반영하여 5가지의 미래토지이용변화 시나리오를 생성하고 이를 유역모형인 SWAT모형에 적용하여 미래기후변화와 토지이용변화를 모두 고려하여 도암댐 유역에서의 유출을 전망하였다.