• Journal of Information Processing Systems
• /
• v.16 no.1
• /
• pp.171-183
• /
• 2020

Capsule endoscopy is one of the increasingly demanded diagnostic methods among patients in recent years because of its ability to observe small intestine difficulties. It is often conducted for 12 to 14 hours, but significant frames constitute only 10% of whole frames. Thus, it has been designed to automatically acquire significant frames through deep learning. For example, studies to track the position of the capsule (stomach, small intestine, etc.) or to extract lesion-related information (polyps, etc.) have been conducted. However, although grouping or labeling the training images according to similar features can improve the performance of a learning model, various attributes (such as degree of wrinkles, presence of valves, etc.) are not considered in conventional approaches. Therefore, we propose a class-labeling method that can be used to design a learning model by constructing a knowledge model focused on main lesions defined in standard terminologies for capsule endoscopy (minimal standard terminology, capsule endoscopy structured terminology). This method enables the designing of a systematic learning model by labeling detailed classes through differentiation of similar characteristics.

• Journal of Information Science Theory and Practice
• /
• v.9 no.4
• /
• pp.35-49
• /
• 2021

In a situation where there are multiple diverse datasets, it is essential to have an efficient method to provide users with the datasets they require. To address this suggestion, necessary datasets should be selected on the basis of the relationships between the datasets. In particular, in order to discover the necessary datasets for disaster resolution, we need to consider the disaster resolution stage. In this paper, in order to provide the necessary datasets for each stage of disaster resolution, we constructed a disaster type and disaster management process ontology and designed a method to determine the necessary datasets for each disaster type and disaster management process step. In addition, we introduce a method to determine relationships between datasets necessary for disaster response. We propose a method for discovering datasets based on minimal relationships such as "isA," "sameAs," and "subclassOf." To discover suitable datasets, we designed a knowledge exploration model and collected 651 disaster-related datasets for improving our method. These datasets were categorized by disaster type from the perspective of disaster management. Categorizing actual datasets into disaster types and disaster management types allows a single dataset to be classified as multiple types in both categories. We built a knowledge exploration model on the basis of disaster examples to ensure the configuration of our model.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.1
• /
• pp.147-165
• /
• 2021

In this work, we empirically evaluated the efficiency of the recent EfficientNetB0 model to identify and diagnose malaria parasite infections in blood smears. The dataset used was collected and classified by relevant experts from the Lister Hill National Centre for Biomedical Communications (LHNCBC). We prepared our samples with minimal image transformations as opposed to others, as we focused more on the feature extraction capability of the EfficientNetB0 baseline model. We applied transfer learning to increase the initial feature sets and reduced the training time to train our model. We then fine-tuned it to work with our proposed layers and re-trained the entire model to learn from our prepared dataset. The highest overall accuracy attained from our evaluated results was 94.70% from fifty epochs and followed by 94.68% within just ten. Additional visualization and analysis using the Gradient-weighted Class Activation Mapping (Grad-CAM) algorithm visualized how effectively our fine-tuned EfficientNetB0 detected infections better than other recent state-of-the-art DCNN models. This study, therefore, concludes that when fine-tuned, the recent EfficientNetB0 will generate highly accurate deep learning solutions for the identification of malaria parasites in blood smears without the need for stringent pre-processing, optimization, or data augmentation of images.

• Journal of the Korea Computer Graphics Society
• /
• v.29 no.1
• /
• pp.23-29
• /
• 2023

In this paper, we introduce a method to obtain high-quality results at a low cost for simulating musculoskeletal characters based on data from the reference motion through motion capture on two-legged walking through reinforcement learning. We reset the motion data of the reference motion to allow the character model to perform, and then train the corresponding motion to be learned through reinforcement learning. We combine motion imitation of the reference model with minimal metabolic energy for the muscles to learn to allow the musculoskeletal model to perform two-legged walking in the desired direction. In this way, the musculoskeletal model can learn at a lower cost than conventional manually designed controllers and perform high-quality bipedal walking.

• The Bulletin of The Korean Astronomical Society
• /
• v.44 no.2
• /
• pp.73.3-73.3
• /
• 2019

The AGN unification model suggested the presence of obscuring material, a dusty torus, to explain the various types of AGN. IR SED model fitting is a crucial tool to probe the structure and properties of the dusty torus. We use a sample of 16 local quasi-stellar objects in Martinez-Paredes et al. (2017) with obtained NIR and MIR high-angular resolution (~0.3") imaging data from EMIR, CIRCE and CanariCam on the 10.4-m Gran Telescopio CANARIAS (GTC) while 4 objects have NIR high-angular resolution photometry from NICMOS/HST from the literature. The unresolved NIR emission from the NIR image analysis and low-resolution Spitzer/IRS spectra are used to construct NIR-MIR SEDs covering a larger spectral range. We investigate the sensitivity of the geometrical (e.g. viewing angle) and physical parameters (e.g. optical depth) of the clumpy dusty torus model of Nenkova et al. and the clumpy disk+outflow model of Hoenig et al. We aim to investigate the minimal dataset needed to well constrain the parameters of the models and derive the properties of the dusty torus. These results will allow us to plan future observations for a larger sample of high luminosity AGNs with the James Webb Space Telescope and the Giant Magellan Telescope.

• Proceedings of the Earthquake Engineering Society of Korea Conference
• /
• 2002.03a
• /
• pp.203-210
• /
• 2002

The past bridge design specifications of Korea didn't include 1imitation on the amount of lap splices in the plastic hinge zone of piers, and so do current specifications. But these specifications include just limitation on the minimal length of lap splices. Thus, a large majority of non-seismically designed bridge piers may have lap splices in plastic hinge zone. In this study, model pier was selected among existent bridge piers whose failure mode is complex shear-flexure mode. Full scaled RC pier models whose aspect ratio is about 2.67 were constructed and quasi static test according to the drift level history was implemented. From the test results, effect of the lap splices on the seismic performance of bridges piers was analyzed, and the seismic capacity of the model bridges was evaluated by capacity spectrum method.

• Journal of the Korean Society of Safety
• /
• v.30 no.4
• /
• pp.174-180
• /
• 2015

This study developed a Safety Culture Impact Assessment Model (SCIAM) which consists of a safety culture assessment methodology and a safety culture impact quantification methodology. The SCIAM uses safety culture impact index (SCII) to monitor the status of safety culture of the NPPs periodically and it uses relative core damage frequency (RCDF) to present the impact of safety culture on the safety of the NPPs. As a result of applying SCIAM to the reference plant (Kori 3), the standard for the healthy safety culture of the reference plant is suggested. SCIAM might contribute to improve the safety of the NPPs (Nuclear Power Plants) by monitoring the status of safety culture periodically and presenting the standard of healthy safety culture.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.7 no.9
• /
• pp.165-171
• /
• 1999

An oil layer fuel absorption /desorption modeling was developed. Multi-component fuel model has showed more reasonable condition than single component model. Henry's constant which is related to solubility is the most important variable in the oil layer absorption/desorption mechanism. The oil segments close to the top of the cylinder liner have more significant contribution to the fuel absorption and desorption process than other oil segments. At the warmed-up condition, the effect of the engine speed on the precent fuel absorbed/desorbed is minimal. But at low il film temperature, percent of fuel abosrbed/desorbed is decreased with increasing the engine speed because of low value of molecular diffusion coefficient of fuel. The amount of fuel trapped in the piston crevice is from 2 to 2.3 times larger than that of fuel in the oil fim. However, fuel form oil film slowly desorbs into the combustion chamber compared with fuel from the piston crevices when the engines is cold.

• International Journal of Quality Innovation
• /
• v.5 no.1
• /
• pp.1-9
• /
• 2004

The nonhomogeneous poisson process (NHPP) is often used to model repairable systems that are subject to a minimal repair strategy, with negligible repair times. In this situation, the system can be characterized by its intensity function. There have been many NHPP models according to intensity functions. However, the intensity function of system in use can be changed because of repair or its aging. We consider the single change point model as the modification of the power law process. The shape parameter of its intensity function is changed before and after the change point. We detect the presence of the change point using Bayesian methodology. Some numerical results are also presented.

• Journal of Korean Society for Quality Management
• /
• v.21 no.2
• /
• pp.109-120
• /
• 1993

Most systems are composed of components which have different failure chracteristics. Since the failure characteristics of components is different, it is rational and reasonable to establish a maintenance model to be considered repair and replacement policies which are proper to failure characteristics of these components. This paper proposes the age replacement time for a system composed of components which have different failure characteristics. In this model, it is assumed that a system is composed of a critical failure component, a major failure component, minor failure component. If any failure occurs to critical component before its age replacement time, the system should be replaced. If any failure does not occur until its age replacement time, preventive replacement should be performed at age replacement time T. Major component is minimal repaired if any failure occurs during operation. Minor component should be replaced as soon as failure is found. This paper determines the optimal replacement time of the system which minimize, total maintenance cost and initial stock Quantity of minor component within this optimal replacement time. Numerical example illustrates these results.