• Journal of the Korea Institute of Military Science and Technology
• /
• v.26 no.5
• /
• pp.392-399
• /
• 2023

This research paper investigates the effectiveness of using computer graphics(CG) based synthetic data for deep learning in military vehicle detection. In particular, we explore the use of synthetic image generation techniques to train deep neural networks for object detection tasks. Our approach involves the generation of a large dataset of synthetic images of military vehicles, which is then used to train a deep learning model. The resulting model is then evaluated on real-world images to measure its effectiveness. Our experimental results show that synthetic training data alone can achieve effective results in object detection. Our findings demonstrate the potential of CG-based synthetic data for deep learning and suggest its value as a tool for training models in a variety of applications, including military vehicle detection.

• Structural Engineering and Mechanics
• /
• v.74 no.1
• /
• pp.55-67
• /
• 2020

The common practice to predict the characteristic structural load effects (LEs) in long reference periods is to employ the extreme value theory (EVT) for building limit distributions. However, most applications ignore that LEs are driven by multiple loading events and thus do not have the identical distribution, a prerequisite for EVT. In this study, we propose the composite extreme value modeling approach using clustering to (a) cluster initial blended samples into finite identical distributed subsamples using the finite mixture model, expectation-maximization algorithm, and the Akaike information criterion; (b) combine limit distributions of subsamples into a composite prediction equation using the generalized Pareto distribution based on a joint threshold. The proposed approach was validated both through numerical examples with known solutions and engineering applications of bridge traffic LEs on a long-span bridge. The results indicate that a joint threshold largely benefits the composite extreme value modeling, many appropriate tail approaching models can be used, and the equation form is simply the sum of the weighted models. In numerical examples, the proposed approach using clustering generated accurate extrema prediction of any reference period compared with the known solutions, whereas the common practice of employing EVT without clustering on the mixture data showed large deviations. Real-world bridge traffic LEs are driven by multi-events and present multipeak distributions, and the proposed approach is more capable of capturing the tendency of tailed LEs than the conventional approach. The proposed approach is expected to have wide applications to general problems such as samples that are driven by multiple events and that do not have the identical distribution.

• Journal of Electrical Engineering and Technology
• /
• v.10 no.2
• /
• pp.699-707
• /
• 2015

Although digit character recognition has got a significant improvement in recent years, it is still challenging to achieve satisfied result if the data contains an amount of distracting factors. This paper proposes a novel digit character recognition approach using a multi-layer hierarchical model, Hybrid Restricted Boltzmann Machines (HRBMs), which allows the learning architecture to be robust to background distracting factors. The insight behind the proposed model is that useful high-level features appear more frequently than distracting factors during learning, thus the high-level features can be decompose into hybrid hierarchical structures by using only small label information. In order to extract robust and compact features, a stochastic 0-1 layer is employed, which enables the model's hidden nodes to independently capture the useful character features during training. Experiments on the variations of Mixed National Institute of Standards and Technology (MNIST) dataset show that improvements of the multi-layer hierarchical model can be achieved by the proposed method. Finally, the paper shows the proposed technique which is used in a real-world application, where it is able to identify digit characters under various complex background images.

• IE interfaces
• /
• v.24 no.2
• /
• pp.128-138
• /
• 2011

This research proposes a resource allocation model of Data QC (Quality Control) activities using COQ (Cost of Quality). The model has been developed based on a series of research efforts such as COQ classifications, weight determination of Data QC activities, and an aggregation approach between COQ and Data QC activities. In the first stage of this research, COQ was divided into the four typical classifications (prevention costs, appraisal costs, internal failure costs and external failure costs) through the opinions from five professionals in Data QC. In the second stage, the weights of Data QC activities were elicited from the field professionals. An aggregation model between COQ and Data QC activities has been then proposed to help the practitioners make a resource allocation strategy. DEA (Data Envelopment Analysis) was utilized for locating efficient decision points. The proposed resource allocation model has been validated using the case of Korea national defense information system. This research is unique in that it applies the concept of COQ to the data management for the first time and that it demonstrates a possible contribution to a real world case for budget allocation of national defense information.

• Structural Monitoring and Maintenance
• /
• v.4 no.3
• /
• pp.195-220
• /
• 2017

Nowadays, railway system plays a significant role in transportation, conveying cargo, passengers, minerals, grains, and so forth. Railway ballasted track is a conventional railway track as can be seen all over the world. Ballast, located underneath the sleepers, is the most important elements on ballasted track, which has many functions and requires routine maintenance. Ballast needs to be maintained frequently to prevent rail buckling, settlement, misalignment so that ballast has to be modelled accurately. Continuum model was introduced to model granular material and was extended in ballast. However, ballast is a heterogeneous material with highly nonlinear behaviour. Hence, ballast could not be modelled accurately in continuum model due to the discontinuities nature and material degradation of ballast. Discrete element modelling (DEM) is proposed as an alternative approach that provides insight into constitutive model, realistic particle, and contact algorithm between each particle. DEM has been studied in many recent decades. However, there are limitations due to the high computational time and memory consumption, which cause the lack of using in high range. This paper presents a review of recent ballast modelling with benefits and drawbacks. Ballast particles are illustrated either circular, circular crump, spherical, spherical crump, super-quadric, polygonal and polyhedral. Moreover, the gaps and limitations of previous studies are also summarized. The outcome of this study will help the understanding into different ballast modelling and particle. The insight information can be used to improve ballast modelling and monitoring for condition-based track maintenance.

• The Journal of Asian Finance, Economics and Business
• /
• v.5 no.4
• /
• pp.95-100
• /
• 2018

The study investigates the factors that affect Kyrgyzstan's bilateral trade flows with its main trading partners and attempts to predict trade potential for Kyrgyzstan. Using panel data, the gravity model is applied to estimate Kyrgyzstan's trade from 2000 to 2016 for its 35 main trading partners. The coefficients derived from the gravity-model estimation are then used to predict trade potential for Kyrgyzstan. Results proved to be successful and explained 63% of the fluctuations in Kyrgyzstan's trade. According to the results, Kyrgyzstan's and its partners' GDP have a positive effect on trade, while distance and partners' population prove to have a negative effect. Predicted trade potential reveals that neighboring countries (China, Kazakhstan, Uzbekistan, and Tajikistan) and Russia still have a significant trade potential. Kyrgyzstan, being a less developed economy, even by Central Asia standards, can only achieve its goals of reducing poverty and becoming more developed by increasing its overall trade with the rest of the world. Therefore, it is essential to study the main determinants of Kyrgyzstan's bilateral trade. In this way, we can help policy makers formulate policies to expand Kyrgyzstan's trade. This study is the first attempt to apply to the gravity model to Kyrgyzstan in an attempt to predict trade potential.

• Journal of the Korean Data and Information Science Society
• /
• v.22 no.2
• /
• pp.197-206
• /
• 2011

The ability to detect online abnormal events in signals is essential in many real-world signal processing applications. In order to detect abnormal events, previously known algorithms require an explicit signal statistical model, and interpret abnormal events as statistical model abrupt changes. In general, maximum likelihood and Bayesian estimation theory to estimate well as detection methods have been used. However, the above-mentioned methods for robust and tractable model, it is not easy to estimate. More freedom to estimate how the model is needed. In this paper, we investigate a machine learning, descriptor-based approach that does not require a explicit descriptors statistical model, based on support vector machines are known to be robust statistical models and a sequential optimal algorithm online support vector machine is introduced.

• Journal of Distribution Science
• /
• v.19 no.8
• /
• pp.25-35
• /
• 2021

Purpose: This study aims to propose an integrated MCDM model to support the qualified personnel selection in the distribution science. Research design, data, and methodology: The integrated approach of AHP and TOPSIS was employed to address the personnel selection problem. The AHP method was used to define the weights of the selection criteria, whereas the TOPSIS was applied to rank alternatives. The proposed model was then applied into a leading logistics company to select the best alternatives to be the sales deputy manager. Results: The results showed that Candidate 3 is the most qualified personnel for the sales deputy manager position as he is ranked first in the order of preference for recruitment. Conclusions: The proposed model provides the decision makers with more effective and time-saving methods than conventional ones. Therefore, the model can be applied to personnel selection around the world. In terms of theoretical contribution, this study proposes a personnel selection model for choosing the most appropriate candidates. In addition, the study adds to the theory of human resources management and logistics management the full set of personnel selection criteria including education, experience, skills, health, personality traits and foreign language.

• The Journal of Asian Finance, Economics and Business
• /
• v.9 no.8
• /
• pp.1-6
• /
• 2022

This paper uses the gravity model estimated by the random effect method to analyze the factors affecting Vietnam's coffee export turnover for the period 2007-2020 major markets according to statistics from the General Statistics Office and the General Department of Customs. Coffee export turnover was collected from the General Statistics Office, General Department of Customs, and Vietnam Cacao Coffee Association. The authors calculated the price of coffee based on output and export value from data on coffee export turnover; the authors calculated the economic gap based on population and Gross Domestic Product data (reference: geographic distance metrics on the website: http://www.distancefromto.net/countries.php) and other data was collected based on the databases of the Food and Agriculture Organization of the United Nations, the International Monetary Fund, and World Bank organizations. The results of the study show that from 2007 to 2020, the factors of Vietnam's export price of coffee, geographical distance, Gross Domestic Product of the importing country and Gross Domestic Product of Vietnam, the population of Vietnam, the economic gap between Vietnam and the importing country, the openness of the economy, all have an impact on Vietnam's coffee export turnover. Finally, some conclusions about the policy's impact are made based on the empirical results of the paper.

• Fourth Industrial Review
• /
• v.3 no.2
• /
• pp.1-10
• /
• 2023

Purpose - The main purpose of the paper is to examine the variables affecting carbon emissions in different nations around the world. Research design, data, and methodology - To measure its impact on carbon emissions, secondary data has data of the top 50 Countries have been taken. The stochastic Impacts by Regression on Population, Affluence, and Technology (STIRPAT) model have been used to quantify the factors that affect carbon emissions. A modified version using Industry 4.0 and region in fundamental STIRPAT model has been applied with the ordinary least square approach. The outcome has been measured using both the basic and extended STIRPAT models. Result - Technology found a positive determinant as well as statistically significant at the alpha level of 0.001models indicating that technological innovation helps reduce carbon emissions. In total, 4 models have been derived to test the best fit and find the highest explaining capacity of variance. Model 3 is found best fit in explanatory power with the highest adjusted R2 (97.95%). Conclusion - It can be concluded that the selected explanatory variables population and Industry 4.0 are found important indicators and causal factors for carbon emission and found constant with all four models for total CO2 and Co2 per capita.