• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.16 no.6
• /
• pp.90-100
• /
• 2017

The purpose of this study is to detect and label the vehicles using the drone images as a way to overcome the limitation of the existing point and section detection system and vehicle gap estimation on Arterial road. In order to select the appropriate time zone, position, and altitude for the acquisition of the drone image data, the final image data was acquired by shooting under various conditions. The vehicle was detected by applying mixed Gaussian, image binarization and morphology among various image analysis techniques, and the vehicle was labeled by applying Kalman filter. As a result of the labeling rate analysis, it was confirmed that the vehicle labeling rate is 65% by detecting 185 out of 285 vehicles. The gap was calculated by pixel unitization, and the results were verified through comparison and analysis with Daum maps. As a result, the gap error was less than 5m and the mean error was 1.67m with the preceding vehicle and 1.1m with the following vehicle. The gaps estimated in this study can be used as the density of the urban roads and the criteria for judging the service level.

• Journal of KIISE
• /
• v.42 no.4
• /
• pp.475-482
• /
• 2015

Developing a high-performance Semantic Role Labeling (SRL) system for a domain requires manually annotated training data of large size in the same domain. However, such SRL training data of sufficient size is available only for a few domains. Performances of Korean SRL are degraded by almost 15% or more, when it is directly applied to another domain with relatively small training data. This paper proposes two techniques to minimize performance degradation in the domain transfer. First, a domain adaptation algorithm for Korean SRL is proposed which is based on the prior model that is one of domain adaptation paradigms. Secondly, we proposed to use simplified features related to morphological and syntactic tags, when using small-sized target domain data to suppress the problem of data sparseness. Other domain adaptation techniques were experimentally compared to our techniques in this paper, where news and Wikipedia were used as the sources and target domains, respectively. It was observed that the highest performance is achieved when our two techniques were applied together. In our system's performance, F1 score of 64.3% was considered to be 2.4~3.1% higher than the methods from other research.

• Journal of the Korea Computer Graphics Society
• /
• v.28 no.3
• /
• pp.55-65
• /
• 2022

Recently, deep learning-based automated systems for identifying and detecting landmarks have been proposed. In order to train such a deep learning-based model without overfitting, a large amount of image and labeling data is required. Conventionally, an experienced reader manually identifies and labels landmarks in a patient's image. However, such measurement is not only expensive, but also has poor reproducibility, so the need for an automated labeling method has been raised. In addition, in the X-ray image, since various human tissues on the path through which the photons pass are displayed, it is difficult to identify the landmark compared to a general natural image or a 3D image modality image. In this study, we propose a geometric data augmentation technique that enables the generation of a large amount of labeling data in X-ray images. In addition, the optimal attention mechanism for landmark detection was presented through the implementation and application of various attention techniques to improve the detection performance of 16 major landmarks in the skull. Finally, among the major cranial landmarks, markers that ensure stable detection are derived, and these markers are expected to have high clinical application potential.

• Journal of KIBIM
• /
• v.13 no.4
• /
• pp.96-105
• /
• 2023

This study focuses on developing a building object recognition technology for efficient use in the remodeling of buildings constructed without drawings. In the era of the 4th industrial revolution, smart technologies are being developed. This research contributes to the architectural field by introducing a deep learning-based method for automatic object classification and recognition, utilizing point cloud data. We use a TD3D network with voxels, optimizing its performance through adjustments in voxel size and number of blocks. This technology enables the classification of building objects such as walls, floors, and roofs from 3D scanning data, labeling them in polygonal forms to minimize boundary ambiguities. However, challenges in object boundary classifications were observed. The model facilitates the automatic classification of non-building objects, thereby reducing manual effort in data matching processes. It also distinguishes between elements to be demolished or retained during remodeling. The study minimized data set loss space by labeling using the extremities of the x, y, and z coordinates. The research aims to enhance the efficiency of building object classification and improve the quality of architectural plans by reducing manpower and time during remodeling. The study aligns with its goal of developing an efficient classification technology. Future work can extend to creating classified objects using parametric tools with polygon-labeled datasets, offering meaningful numerical analysis for remodeling processes. Continued research in this direction is anticipated to significantly advance the efficiency of building remodeling techniques.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.19 no.1
• /
• pp.189-196
• /
• 2024

Frequent false positives alarm from the Intelligent Selective Control System have raised significant concerns. These persistent issues have led to declines in operational efficiency and market credibility among agents. Developing a new model or replacing the existing one to mitigate false positives alarm entails substantial opportunity costs; hence, improving the quality of the training dataset is pragmatic. However, smaller organizations face challenges with inadequate capabilities in dataset collection and refinement. This paper proposes an automatic human pose data collection system centered around a human pose estimation model, utilizing camera-based sensor fusion techniques and edge devices. The system facilitates the direct collection and real-time processing of field data at the network periphery, distributing the computational load that typically centralizes. Additionally, by directly labeling field data, it aids in constructing new training datasets.

• The Research Journal of the Costume Culture
• /
• v.17 no.2
• /
• pp.203-215
• /
• 2009

The purpose of this study was to investigate current garment sizes of women's formal jackets, targeting plus-size women in online shopping malls, and to identify effective size information involved in online apparel purchase behaviors to overcome the short comings of current garment sizes from the perspectives of consumers. Basic 88 size formal jackets from the seven companies found on the 22 websites were collected and analyzed. The data were collected from March to October 2007, and analyzed using SPSS 14.0. The results were summarized as follows. First, there was no website using standard garment size labeling with 'bust-hip-height' set up by KS K 0051 among the 22 websites. Instead, all 22 websites used garment size labeling with figures such as 88, 99, 100, 110, and 120 or with letters such as L, XL, and XXL. The websites presented no body size, but listed garment size. Furthermore, the size information was presented differently, ranging from three items of bust circumference, upper arm length, and jacket length to six items of shoulder width, bust circumference, waist circumference, sleeve width, sleeve length, and jacket length. In addition, no website presented basic information for hip circumference, despite the jacket length covering the hips. Second, a total of 85.7% the websites listed bust circumferences in 88 garment sizes collected as 100cm. Shoulder widths were presented as 39cm or 37cm. Sleeve circumferences were addressed the same, 36cm, in all websites. Third, comparing the differences between guidance sizes and measurement sizes, only 28.5% of the web sites posted guidance sizes of shoulder widths the same as those of the measurement sizes. All web sites presented guidance sizes of bust circumstances as 1 to 5cm smaller than those of the measurement sizes.

• Korean Journal of Radiology
• /
• v.22 no.9
• /
• pp.1537-1546
• /
• 2021

Objective: To assess the role of arterial spin-labeling (ASL) perfusion MRI in identifying cerebral perfusion changes after indirect revascularization in children with moyamoya disease. Materials and Methods: We included pre- and postoperative perfusion MRI data of 30 children with moyamoya disease (13 boys and 17 girls; mean age ± standard deviation, 6.3± 3.0 years) who underwent indirect revascularization between June 2016 and August 2017. Relative cerebral blood flow (rCBF) and qualitative perfusion scores for arterial transit time (ATT) effects were evaluated in the middle cerebral artery (MCA) territory on ASL perfusion MRI. The rCBF and relative time-to-peak (rTTP) values were also measured using dynamic susceptibility contrast (DSC) perfusion MRI. Each perfusion change on ASL and DSC perfusion MRI was analyzed using the paired t test. We analyzed the correlation between perfusion changes on ASL and DSC images using Spearman's correlation coefficient. Results: The ASL rCBF values improved at both the ganglionic and supraganglionic levels of the MCA territory after surgery (p = 0.040 and p = 0.003, respectively). The ATT perfusion scores also improved at both levels (p < 0.001 and p < 0.001, respectively). The rCBF and rTTP values on DSC MRI showed significant improvement at both levels of the MCA territory of the operated side (all p < 0.05). There was no significant correlation between the improvements in rCBF values on the two perfusion images (r = 0.195, p = 0.303); however, there was a correlation between the change in perfusion scores on ASL and rTTP on DSC MRI (r = 0.701, p < 0.001). Conclusion: Recognizing the effects of ATT on ASL perfusion MRI may help monitor cerebral perfusion changes and complement quantitative rCBF assessment using ASL perfusion MRI in patients with moyamoya disease after indirect revascularization.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.14 no.4
• /
• pp.313-321
• /
• 2014

In the current era of data-intensive services, the handling of big data is a crucial issue that affects almost every discipline and industry. In this study, we propose a classification method for large volumes of numeric data, which is implemented in a distributed programming framework, i.e., MapReduce. The proposed method partitions the data space into a grid structure and it then models the probability distributions of classes for grid cells by collecting sufficient statistics using distributed MapReduce tasks. The class labeling of new data is achieved by k-nearest neighbor classification based on Bayesian inference.

• IEIE Transactions on Smart Processing and Computing
• /
• v.5 no.3
• /
• pp.199-206
• /
• 2016

This paper describes a simple routing control system. We propose achieving high-speed data transmission without modifying the data frame configuration. To add a routing control signal, called the "labeling signal" in this paper, to the data frame, we use a frequency modulation technique on the transmitted frame. This means you need not change the data frame when you transmit additional signals. Using a prototype system comprising a field-programmable gate array and discrete elements, we investigate the system performance and devise a method to achieve high resolution. A three-channel routing control for a 10 Gb/s data frame was achieved, which confirms the advantages of the proposed system.

• The Korean Journal of Nuclear Medicine
• /
• v.30 no.3
• /
• pp.361-366
• /
• 1996

This study was to determine the effect of $Al^{3+}$ in $^{99m}Tc$ eluate from $^{99}Mo-^{99m}Tc$ generator on labeling efficiency and biodistribution of $^{99m}Tc$-MDP. The chromatographic analysis of $^{99m}Tc$-MDP preparations containing $Al^{3+}(0-62.5{\mu}g/ml)$ showed decreased labeling efficiency $^{99m}Tc$ pertechnetate and hydrolyzed reduced $^{99m}Tc$ fraction increased with increasing concentrations of aluminum. However, the chromatography system could not discern between hydrolyzed reduced $^{99m}Tc$ and $^{99m}Tc$ labeled colloid. $^{99m}Tc$-MDP preparations containing aluminum were relatively stable. Chromatographic analysis also confirmed that no significant differences were observed in the radiochemical purity of the filtered and the unfiltered $^{99m}Tc$-MDP preparations containing aluminum by $0.22{\mu}m$ syringe filter. In biodistribution data of ICR-mice, blood and heart uptake were increasing with increasing concentrations of aluminum, because of decreasing labeling efficiency of $^{99m}Tc$-MDP and increasing of $^{99m}Tc$ pertechnetate. However, liver and bone uptake were not significantly increased. In rat images no difference were observed at $5{\mu}g/ml\;Al^{3+}$ compare with at $0{\mu}g/ml\;Al^{3+}$, but at $10{\mu}g/ml\;Al^{3+}$ lumbar uptake was increased. As a practical conclusion, a concentration below $10{\mu}g/ml\;Al^{3+}$($10{\mu}g/ml\;Al^{3+}$ is the maximum allowed in pertechnetate eluate from $^{99}Mo-^{99m}Tc$ generator by USP.) in $^{99m}Tc$-MDP radiopharmaceutical result in low labeling efficiency. Radiochemical purity 90% of $^{99m}Tc$-MDP is the minimum allowed by USP. Therefore, when soft tissue uptake is observed in $^{99m}Tc$-MDP bone scan and labeling efficiency is above 90%, we can expect that $Al^{3+}$ in pertechnetated eluate is not the cause of soft tissue uptake.