• Korean Journal of Remote Sensing
• /
• v.38 no.2
• /
• pp.199-213
• /
• 2022

The unsupervised domain adaptation can solve the impractical issue of repeatedly collecting high-quality training data every year for annual crop classification. This study evaluates the applicability of deep learning-based unsupervised domain adaptation models for crop classification. Three unsupervised domain adaptation models including a deep adaptation network (DAN), a deep reconstruction-classification network, and a domain adversarial neural network (DANN) are quantitatively compared via a crop classification experiment using unmanned aerial vehicle images in Hapcheon-gun and Changnyeong-gun, the major garlic and onion cultivation areas in Korea. As source baseline and target baseline models, convolutional neural networks (CNNs) are additionally applied to evaluate the classification performance of the unsupervised domain adaptation models. The three unsupervised domain adaptation models outperformed the source baseline CNN, but the different classification performances were observed depending on the degree of inconsistency between data distributions in source and target images. The classification accuracy of DAN was higher than that of the other two models when the inconsistency between source and target images was low, whereas DANN has the best classification performance when the inconsistency between source and target images was high. Therefore, the extent to which data distributions of the source and target images match should be considered to select the best unsupervised domain adaptation model to generate reliable classification results.

• Korean Journal of Exercise Nutrition
• /
• v.16 no.2
• /
• pp.65-71
• /
• 2012

Oxygen is the final acceptor of electron transport from fat and carbohydrate oxidation, which is the rate-limiting factor for cellular ATP production. Under altitude hypoxia condition, energy reliance on anaerobic glycolysis increases to compensate for the shortfall caused by reduced fatty acid oxidation [1]. Therefore, training at altitude is expected to strongly influence the human metabolic system, and has the potential to be designed as a non-pharmacological or recreational intervention regimen for correcting diabetes or related metabolic problems. However, most people cannot accommodate high altitude exposure above 4500 M due to acute mountain sickness (AMS) and insulin resistance corresponding to a increased levels of the stress hormones cortisol and catecholamine [2]. Thus, less stringent conditions were evaluated to determine whether glucose tolerance and insulin sensitivity could be improved by moderate altitude exposure (below 4000 M). In 2003, we and another group in Austria reported that short-term moderate altitude exposure plus endurance-related physical activity significantly improves glucose tolerance (not fasting glucose) in humans [3,4], which is associated with the improvement in the whole-body insulin sensitivity [5]. With daily hiking at an altitude of approximately 4000 M, glucose tolerance can still be improved but fasting glucose was slightly elevated. Individuals vary widely in their response to altitude challenge. In particular, the improvement in glucose tolerance and insulin sensitivity by prolonged altitude hiking activity is not apparent in those individuals with low baseline DHEA-S concentration [6]. In addition, hematopoietic adaptation against altitude hypoxia can also be impaired in individuals with low DHEA-S. In short-lived mammals like rodents, the DHEA-S level is barely detectable since their adrenal cortex does not appear to produce this steroid [7]. In this model, exercise training recovery under prolonged hypoxia exposure (14-15% oxygen, 8 h per day for 6 weeks) can still improve insulin sensitivity, secondary to an effective suppression of adiposity [8]. Genetically obese rats exhibit hyperinsulinemia (sign of insulin resistance) with up-regulated baseline levels of AMP-activated protein kinase and AS160 phosphorylation in skeletal muscle compared to lean rats. After prolonged hypoxia training, this abnormality can be reversed concomitant with an approximately 50% increase in GLUT4 protein expression. Additionally, prolonged moderate hypoxia training results in decreased diffusion distance of muscle fiber (reduced cross-sectional area) without affecting muscle weight. In humans, moderate hypoxia increases postprandial blood distribution towards skeletal muscle during a training recovery. This physiological response plays a role in the redistribution of fuel storage among important energy storage sites and may explain its potent effect on changing body composition. Conclusion: Prolonged moderate altitude hypoxia (rangingfrom 1700 to 2400 M), but not acute high attitude hypoxia (above 4000 M), can effectively improve insulin sensitivity and glucose tolerance for humans and antagonizes the obese phenotype in animals with a genetic defect. In humans, the magnitude of the improvementvaries widely and correlates with baseline plasma DHEA-S levels. Compared to training at sea-level, training at altitude effectively decreases fat mass in parallel with increased muscle mass. This change may be associated with increased perfusion of insulin and fuel towards skeletal muscle that favors muscle competing postprandial fuel in circulation against adipose tissues.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.22 no.2
• /
• pp.261-269
• /
• 2018

This paper presents a convenient method to estimate the internal parameters of plenoptic camera using CCD(charge-coupled device) camera model. The images used for plenoptic camera calibration generally use the checkerboard pattern used in CCD camera calibration. Based on the CCD camera model, the determinant of the plenoptic camera model can be derived through the relationship with the plenoptic camera model. We formulate four equations that express the focal length, the principal point, the baseline, and distance between the virtual camera and the object. By performing a nonlinear optimization technique, we solve the equations to estimate the parameters. We compare the estimation results with the actual parameters and evaluate the reprojection error. Experimental results show that the MSE(mean square error) is 0.309 and estimation values are very close to actual values.

• Journal of the Korean Data and Information Science Society
• /
• v.20 no.2
• /
• pp.377-385
• /
• 2009

Twenty five female students in Seoul participated and were divided into two group in the experiment of weight loss effect of two treatments. Fourteen students(Treatment A group), randomly chosen from the students, had fed on diet foods and exercised over 8 weeks, and the remaining students(Treatment B group) had fed on diet foods only for the same periods. Weights of 25 students had been measured repeatedly four times at an interval of two weeks during 8 weeks, It resulted from mixed model analysis of repeated measurements data that separate Toeplitz pattern for each treatment group was selected as the optimal covariance pattern. Based upon the optimal covariance pattern model, the baseline effect and time effect were found to be highly significant, but the treatment-time interaction effect was found to be insignificant. Finally, the students with diet foods and exercises were more effective in losing weight than the students with only diet foods were.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.17 no.1
• /
• pp.31-41
• /
• 2018

This study aims to analyze region-based pedestrian evacuation behaviors and information offering effect using Social Force Model, which is micro simulation. All pedestrians were assumed to move to shelters through pedestrian roads according to guidance information at emergency situations, and the pedestrians were classified into adults and the handicapped. According to the results of the road network analysis and simulation analysis, the shelters to which pedestrians can move within the shortest time from each zone were selected as optimum shelters. From this study, the analysis showed that the information provision effects are informative even though total evacuation time increases due to the increase of pedestrian conflict. This study can be used as baseline data for urban area's pedestrian disaster prevention plans.

• Human Ecology Research
• /
• v.58 no.3
• /
• pp.333-341
• /
• 2020

This study examined the trajectories of the marital satisfaction among middle and old aged couples for 8 years. This study focused on middle and old aged couples who were relatively uninterested in family research. Using the 1-5 Waves of Korean Longitudinal Study of Ageing (KLoSA) 2006-2014, the study included 635 middle and old aged couples whose husbands were 65-74 years old at baseline. First, a latent growth model indicated that a husband's and wife's marital satisfaction trajectories decreased over time. Second, a parallel latent growth model indicated that a husband's marital satisfaction trajectories decreased more quickly, when the initial level of their own marital satisfaction was higher (husband's actor-effect); in addition, a wife's marital satisfaction trajectories also decreased more quickly, when the initial level of their own marital satisfaction was higher (wife's actor-effect). Wife's marital satisfaction trajectories decreased more slowly, when the initial level of their husband's marital satisfaction was higher (husband's partner-effect); however, the husband's slope was not associated with the initial level of their wife's. The methodological value of this work shows the application of a parallel latent growth model that includes the actor and partner effect of APIM. The findings extend our understanding of longitudinal marital dynamics between couples and promote a study on the trajectories of marital satisfaction at stages of the family life cycle.

• The Journal of Industrial Distribution & Business
• /
• v.11 no.3
• /
• pp.39-50
• /
• 2020

Purpose: Agriculture, which is heavily influenced by climate conditions, is one of the industries most affected by climate change. In this respect, various studies on the impact of climate change on the agricultural market have been conducted. Since climate change is a long-term phenomenon for more than a decade, long-term projections of agricultural prices as well as climate variables are needed to properly analyze the impact of climate change on the agricultural market. However, these long-term price projections are often major constraints on studies of climate changes. The purpose of this study is to analyze the impacts of climate changes on the Korean onion market using ex-post analysis approach in order to avoid the difficulties of long-term price projections. Research design, data and methodology: This study develops an annual dynamic partial equilibrium model of Korean onion market. The behavioral equations of the model were estimated by OLS based on the annual data from 1988 to 2018. The modelling system is first simulated to have actual onion market conditions from 2014 to 2018 as a baseline and then compared it to the scenario assuming the climatic conditions under RCP8.5 over the same period. Scenario analyses were simulated by both comparative static and dynamic approach to evaluate the differences between the two approaches. Results: According to the empirical results, if the climate conditions under RCP8.5 were applied from 2014 to 2018, the yield of onion would increase by about 4%, and the price of onion would decrease from 3.7% to 17.4%. In addition, the average price fluctuation rate over the five years under RCP8.5 climate conditions is 56%, which is more volatile than 46% under actual climate conditions. Empirical results also show that the price decreases have been alleviated in dynamic model compared with comparative static model. Conclusions: Empirical results show that climate change is expected to increase onion yields and reduce onion prices. Therefore, the appropriate countermeasures against climate change in Korean onion market should be found in the stabilization of supply and demand for price stabilization rather than technical aspects such as the development of new varieties to increase productivity.

• The KIPS Transactions:PartB
• /
• v.16B no.3
• /
• pp.247-256
• /
• 2009

Typical pseudo-relevance feedback methods assume the top-retrieved documents are relevant and use these pseudo-relevant documents to expand terms. The initial retrieval set can, however, contain a great deal of noise. In this paper, we present a cluster-based resampling method to select better pseudo-relevant documents based on the relevance model. The main idea is to use document clusters to find dominant documents for the initial retrieval set, and to repeatedly feed the documents to emphasize the core topics of a query. Experimental results on large-scale web TREC collections show significant improvements over the relevance model. For justification of the resampling approach, we examine relevance density of feedback documents. The resampling approach shows higher relevance density than the baseline relevance model on all collections, resulting in better retrieval accuracy in pseudo-relevance feedback. This result indicates that the proposed method is effective for pseudo-relevance feedback.

• The Journal of the Korea Contents Association
• /
• v.9 no.3
• /
• pp.9-19
• /
• 2009

In this paper, we propose a novel hardware architecture to facilitate the applicable SoC chip design of H.264/SVC which has a great deal of advancement in the international standardization in recent. Moreover, a new C-model simulator based on the proposed hardware system will be presented to support optimal SoC circuit development. Since the proposed SVC decoder is consist of some hardware engine for processing of major decoding tools and core processor for software processing, the system is simply implemented with the conventional embedded system. To improve the feasibility and applicability, and reduce the decoder complexity, the hardware decoder architecture is constructed with only the consideration of IPPP structure scalability without using the full B-picture. Finally, we present results of decoder hardware implementation and decoded picture to show the effectiveness of the proposed hardware architecture and C-model simulator.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.16 no.9
• /
• pp.1605-1614
• /
• 1992

The effects of % overload (% O.L), baseline stress intensity factor range (.DELTA. $K_{b}$) and dimension-less crack depth (a/W) are examined for the retardation behaviors after a single overload and high-low block loads in 7075-T73 aluminum alloy. And wheeler model, which is one of the fatigue life prediction models, is modified to predict retardation life using these test results. The retardation cycles( $N_{d}$) increased with a decrease in a/W and an increase in % O.L. and (.DELTA. $K_{b}$) These effects are more severe after high-low block loads than single overload. In the case of single overload, the main mechanisms of the retardation are the crack closure and the relaxation of K due to crack branching. But in the case of high-low block loads, that of the main mechanism is the crack closure caused by the accumulated compressive residual stree at the crack tip, which is related with the contact of fracture surfaces. Test results were multiple regression analyzed and got regressed shaping correction factors, (n)$_{REG}$, as function of %O.L., a/W and (.DELTA. $K_{b}$) Wheeler model is modified by using these (n)$_{REG}$. The number of delay cycles calculated by modified Wheeler model were in good agreement with the test results of this study.y.udy.y.y.y.