• Journal of Internet Computing and Services
• /
• v.22 no.5
• /
• pp.47-55
• /
• 2021

Analysis of research trends in specific subject areas is performed by examining related topics and subject changes by using topic modeling techniques through keyword extraction for most of the literature information (paper, patents, etc.). Unlike existing research methods, this paper extracts topics related to the research topic using the LDA topic modeling technique for the project information of national R&D projects provided by the National Science and Technology Knowledge Information Service (NTIS) in the field of artificial intelligence. By analyzing these topics, this study aims to analyze research topics and investment directions for national R&D projects. NTIS provides a vast amount of national R&D information, from information on tasks carried out through national R&D projects to research results (thesis, patents, etc.) generated through research. In this paper, the search results were confirmed by performing artificial intelligence keywords and related classification searches in NTIS integrated search, and basic data was constructed by downloading the latest three-year project information. Using the LDA topic modeling library provided by Python, related topics and keywords were extracted and analyzed for basic data (research goals, research content, expected effects, keywords, etc.) to derive insights on the direction of research investment.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.9
• /
• pp.375-384
• /
• 2021

Deep learning techniques have been proven to have high performance in image processing and are applied in various fields. The most widely used methods for validating a deep learning model include a holdout verification method, a k-fold cross verification method, and a bootstrap method. These legacy methods consider the balance of the ratio between classes in the process of dividing the data set, but do not consider the ratio of various features that exist within the same class. If these features are not considered, verification results may be biased toward some features. Therefore, we propose a deep learning model validation method based on data feature coverage for image classification by improving the legacy methods. The proposed technique proposes a data feature coverage that can be measured numerically how much the training data set for training and validation of the deep learning model and the evaluation data set reflects the features of the entire data set. In this method, the data set can be divided by ensuring coverage to include all features of the entire data set, and the evaluation result of the model can be analyzed in units of feature clusters. As a result, by providing feature cluster information for the evaluation result of the trained model, feature information of data that affects the trained model can be provided.

• Journal of Korea Water Resources Association
• /
• v.51 no.12
• /
• pp.1181-1194
• /
• 2018

This study suggests a novel approach of estimating stream flow discharge using the Synthetic Aperture Radar (SAR) images taken from 2015 to 2017 by European Space Agency Sentinel-1 satellite. Fifteen small to medium sized rivers in the Han River basin were selected as study area, and the SAR satellite images and flow data from water level and flow observation system operated by the Korea Institute of Hydrological Survey were used for model construction. First, we apply the histogram matching technique to 12 SAR images that have undergone various preprocessing processes for error correction to make the brightness distribution of the images the same. Then, the flow estimation model was constructed by deriving the relationship between the area of the stream water body extracted using the threshold classification method and the in-situ flow data. As a result, we could construct a power function type flow estimation model at the fourteen study areas except for one station. The minimum, the mean, and the maximum coefficient of determination ($R^2$) of the models of at fourteen study areas were 0.30, 0.80, and 0.99, respectively.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.30 no.6
• /
• pp.507-517
• /
• 2019

Multi-function radar(MFR) is a system that uses various functions such as detection, tracking, and classification. To operate the functions in real-time, the detection stage in MFR usually uses radar signals for short measurement time. We can utilize several conventional detectors in the MFR system to detect low radar cross section maritime targets in the sea-clutter; however, the detectors, which have been developed to be effective for radar signals measured for a longer time, may be inappropriate for MFR. In this study, we proposed a modelling technique of sea-clutter short measurement time. We combined the modeled sea-clutter signal with the maritime-target signal, which was obtained by the numerical analysis method. Using this combined model, we exploited four independent detectors and analyzed the detection performances.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2014.10a
• /
• pp.271-274
• /
• 2014

This paper proposes high-performance SAO(Sample Adaptive Offset) in HEVC(High Efficiency Video Coding) encoder for Ultra HD video processing in real time. SAO is a newly adopted technique belonging to the in-loop filter in HEVC. The proposed SAO encoder hardware architecture uses three-layered buffers to minimize memory access time and to simplify pixel processing and also uses only adder, subtractor, shift register and feed-back comparator to reduce area. Furthermore, the proposed architecture consists of pipelined pixel classification and applying SAO parameters, and also classifies four consecutive pixels into EO and BO concurrently. These result in the reduction of processing time and computation. The proposed SAO encoder architecture is designed by Verilog HDL, and implemented by 180k logic gates in TSMC $0.18{\mu}m$ process. At 110MHz, the proposed SAO encoder can support 4K Ultra HD video encoding at 30fps in real time.

• Korean Journal of Heritage: History & Science
• /
• v.33
• /
• pp.210-255
• /
• 2000

This study aims to classify the framework and arrangement of interior columns (Naeju) which are used in single-story Buddhist halls into several types, and to develop a theory on the process of changes among those types. Since interior columns are building materials which hold up the roof structure and make partitions in the interior space of halls, their framework and arrangement is closely linked to the development of building technology and is expected to reflect new architectural needs. The kinds of interior columns classified by the shape of framework are goju, chaduju, oepyonju, naepyonju. The arrangement of interior columns can he classified by two methods: One which counts the number of the interior column arrangements in a hall, and the other whose classification relates with the side wall columns - Jeongchibup and yijubup. With the combination of these classifications, we can divide the framework and arrangement of interior columns into 8 types From the remains of Korean and Chinese Architecture, we can presume that before the late-Goryo period, jeongchibup had always been applied in the construction of Buddhist halls, and gamju(column reducing) had only been used in examples of small scale. After the founding of Choseon Kingdom, however, national policy had weakened the economic power of Buddhist temples. Because of that, large-scale outdoor Buddhist mass was replaced by small-scale indoor mass, and for this reason, though the scale of Buddhist halls became smaller, the need for a broad interior space became stronger. Thus in early-Choseon period, reduction of interior columns became widely spread. Those types of framework and arrangement of interior columns where yijubup was applied were developed because the rear interior columns arrangements, in order to expand the interior space, have moved backward. Among these types, yiju-goju and yiju-chaduju were developed for the Buddhist halls with paljak roof(hipped-gabled roof), where the load of their side eaves caused structural problems at the side walls. And oepyonju type was for the small-scale and middle-scale Buddhist halls which needed more interior space but didn't want the extension of roof structure. From the local and periodic distribution of each types, we can conclude that the types jeongchi-goju, jeongchi-chaduju and yiju-chaduju have been settled as typical technique of local carpenters. Oepyonju was developed later than the other types, but for its merit of low cost, it became a popular type across the nation.

• Journal of Conservation Science
• /
• v.37 no.3
• /
• pp.232-244
• /
• 2021

This study examined the physical attributes and heat treatment characteristics of glass beads excavated from the Gimhae area, which is the location of Geumgwan Gaya. This enabled classification of surface characteristics of the beads based on the investigation of the color, size, and shape. The glass beads were classified into eight color systems, with purplish-blue beads as the representative color. Bead size was categorized into three types depending on the outer diameter and how it increased over time. Bead shapes were categorized as round, tubular, or doughnut-shaped based on the inner diameter and length, with round being the typical shape. According to the degree of heat treatment, there are three types of cross-section for glass beads that are manufactured by the drawing technique, most of which are the HT-III type. In addition, it is estimated that the heat treatment technology has more considerable effects than other methods. Through non-destructive analysis, the chemical composition was obtained and categorized as flux, stabilizer, and colorant. Analysis confirmed the presence of 63 and 9 pieces in the potash and soda glass groups, respectively. Overall findings from the study highlighted a correlation between the chemical composition and the external factors such as color, size, shape, and manufacturing technology of glass beads recovered from Geumgwan Gaya, revealing characteristics related to that time and region.

• Journal of Korean Neurosurgical Society
• /
• v.64 no.5
• /
• pp.751-762
• /
• 2021

Objective : Endovascular treatment of intracranial aneurysms is challenging in case of wide-necked aneurysms because coils are prone to herniate into the parent artery, causing thromboembolic events or vessel occlusion. This study aims to compare long-term angiographic results of wide-necked aneurysms treated by stent-assisted, double-microcatheter, or single-microcatheter groups. Methods : Between January 2003 and October 2016, 108 aneurysms that were treated with endovascular coil embolization with a neck size wider than 4 mm and a follow-up period of more than 3 years were selected. We performed coil embolization with single-microcatheter, double-microcatheter, and stent-assisted techniques. Angiographic results were evaluated using the Raymond-Roy occlusion classification (RROC). All medical and angiographic records were reviewed retrospectively. Results : Clinical and angiographic analyses were conducted in 108 wide-necked aneurysms. The immediate post-procedural results revealed RROC class I (complete occlusion) in 66 cases (61.1%), class II (residual neck) in 36 cases (33.3%), and class III (residual sac) in six cases (5.6%). The final follow-up results revealed class I in 48 cases (44.4%), class II in 49 cases (45.4%), and class III in 11 cases (10.2%). Of a total of 45 (41.6%) radiologic recurrences, there were 21 cases (19.4%) of major recurrence that required additional treatment, and 24 cases (22.2%) of minor recurrence. The final follow-up angiographic results showed statistically significant differences between the stent-assisted group and the others (p<0.01). Conclusion : Long-term follow-up angiography demonstrated that the stent-assisted technique had a better complete occlusion rate than the other two techniques.

• The Journal of the Convergence on Culture Technology
• /
• v.7 no.3
• /
• pp.577-582
• /
• 2021

Recently, with the development of AR, VR, and smart device technologies, the demand for services based on non-face-to-face environments is also increasing in the fitness industry. The non-face-to-face online home training service has the advantage of not being limited by time and place compared to the existing offline service. However, there are disadvantages including the absence of exercise equipment, difficulty in measuring the amount of exercise and chekcing whether the user maintains an accurate exercise posture or not. In this study, we develop a standard exercise program that can compensate for these shortcomings and propose a new non-face-to-face home training application by using a deep learning-based body posture estimation image processing algorithm. This application allows the user to directly watch and follow the trainer of the standard exercise program video, correct the user's own posture, and perform an accurate exercise. Furthermore, if the results of this study are customized according to their purpose, it will be possible to apply them to performances, films, club activities, and conferences

• Journal of Intelligence and Information Systems
• /
• v.28 no.2
• /
• pp.307-332
• /
• 2022

One of the most intensively conducted research areas in business application study is a bankruptcy prediction model, a representative classification problem related to loan lending, investment decision making, and profitability to financial institutions. Many research demonstrated outstanding performance for bankruptcy prediction models using artificial intelligence techniques. However, since most machine learning algorithms are "black-box," AI has been identified as a prominent research topic for providing users with an explanation. Although there are many different approaches for explanations, this study focuses on explaining a bankruptcy prediction model using a counterfactual example. Users can obtain desired output from the model by using a counterfactual-based explanation, which provides an alternative case. This study introduces a counterfactual generation technique based on a genetic algorithm (GA) that leverages both domain knowledge (i.e., causal feasibility) and feature importance from a black-box model along with other critical counterfactual variables, including proximity, distribution, and sparsity. The proposed method was evaluated quantitatively and qualitatively to measure the quality and the validity.