• International journal of advanced smart convergence
• /
• v.12 no.4
• /
• pp.134-141
• /
• 2023

We present a novel method aimed at refining ground truth data through regularization and modification, particularly applicable when working with the original ground truth set. Enhancing the performance of deep neural networks is achieved by applying regularization techniques to the existing ground truth data. In many machine learning tasks requiring pixel-level segmentation sets, accurately delineating objects is vital. However, it proves challenging for thin and elongated objects such as blood vessels in X-ray coronary angiography, often resulting in inconsistent generation of ground truth data. This method involves an analysis of the quality of training set pairs - comprising images and ground truth data - to automatically regulate and modify the boundaries of ground truth segmentation. Employing the active contour model and a recursive ground truth generation approach results in stable and precisely defined boundary contours. Following the regularization and adjustment of the ground truth set, there is a substantial improvement in the performance of deep neural networks.

• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.10
• /
• pp.409-418
• /
• 2022

In this paper, we propose SAAnnot-C3Pap, a semi-automatic annotation method for obtaining ground truth of a player's posture. In order to obtain ground truth about the two-dimensional joint position in the existing music domain, openpose, a two-dimensional posture estimation method, was used or manually labeled. However, automatic annotation methods such as the existing openpose have the disadvantages of showing inaccurate results even though they are fast. Therefore, this paper proposes SAAnnot-C3Pap, a semi-automated annotation method that is a compromise between the two. The proposed approach consists of three main steps: extracting postures using openpose, correcting the parts with errors among the extracted parts using supervisely, and then analyzing the results of openpose and supervisely. Perform the synchronization process. Through the proposed method, it was possible to correct the incorrect 2D joint position detection result that occurred in the openpose, solve the problem of detecting two or more people, and obtain the ground truth in the playing posture. In the experiment, we compare and analyze the results of the semi-automated annotation method openpose and the SAAnnot-C3Pap proposed in this paper. As a result of comparison, the proposed method showed improvement of posture information incorrectly collected through openpose.

• Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference
• /
• 2006.04a
• /
• pp.99-104
• /
• 2006

The new gravity field combination models are expected to improve the knowledge of the Earth's global gravity field. This study evaluates eleven global gravity field models derived from gravimetry and altimetry surface data in a comparison with ground truth in South Korea. Geoid heights obtained from GPS and levelling in South Korea are compared with geoid heights from the models. The results show that the gravity satellites CHAMP, GRACE and LAGEOS plus gravimetry and altimetry surface data have led to an improvement in gravity field models. As expected, the new combination gravity field model which are EIGEN-CG03C and EIGEN-GL04C give better results than the predecessors widely used models(EGM96, OSU91A etc.).

• Journal of Korea Multimedia Society
• /
• v.20 no.5
• /
• pp.769-781
• /
• 2017

Deep learning networks like Convolutional Neural Networks (CNNs) show successful performances in many computer vision applications such as image classification, object detection, and so on. For implementation of deep learning networks in embedded system with limited processing power and memory, deep learning network may need to be simplified. However, simplified deep learning network cannot learn every possible scene. One realistic strategy for embedded deep learning network is to construct a simplified deep learning network model optimized for the scene images of the installation place. Then, automatic training will be necessitated for commercialization. In this paper, as an intermediate step toward automatic training under fisheye camera environments, we study more precise human localization in fisheye images, and propose an accurate human localization method, Automatic Ground-Truth Labelling Method (AGTLM). AGTLM first localizes candidate human object bounding boxes by utilizing GoogLeNet-LSTM approach, and after reassurance process by GoogLeNet-based CNN network, finally refines them more correctly and precisely(tightly) by applying saliency object detection technique. The performance improvement of the proposed human localization method, AGTLM with respect to accuracy and tightness is shown through several experiments.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.16 no.6
• /
• pp.323-330
• /
• 2021

Visual object tracking is a challenging area of study in the field of computer vision due to many difficult problems, including a fast variation of target shape, occlusion, and arbitrary ground truth object designation. In this paper, we focus on the reinforced feature of the dynamic search area to get better performance than conventional discriminative model prediction trackers on the condition when the accuracy deteriorates since low feature discrimination. We propose a reinforced input feature method shown like the spotlight effect on the dynamic search area of the target tracking. This method can be used to improve performances for deep learning based discriminative model prediction tracker, also various types of trackers which are used to infer the center of the target based on the visual object tracking. The proposed method shows the improved tracking performance than the baseline trackers, achieving a relative gain of 38% quantitative improvement from 0.433 to 0.601 F-score at the visual object tracking evaluation.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.22 no.3
• /
• pp.131-146
• /
• 2023

Smart card data is widely used in the public transportation field. Despite the inevitability of various errors occur during the data collection and storage; however, smart card data errors have not been extensively studied. This paper investigates inherent errors in boarding and alighting station information in smart card data. A comparison smart card data and bus boarding and alighting survey data for the same time frame shows that boarding station names differ by 6.2% between the two data sets. This indicates that the error rate of smart card data is 6.2% in terms of boarding station information, given that bus boarding and alighting survey data can be considered as ground truth. This paper propose 6-step algorithm for correcting errors in smart card boarding station information, linking them to corresponding information in Bus Information System(BIS) Data. Comparing BIS data and bus boarding and alighting survey data for the same time frame reveals that boarding station names correspond by 98.3% between the two data sets, indicating that BIS data can be used as reliable reference for ground truth. To evaluate its performance, applying the 6-step algorithm proposed in this paper to smart card data set shows that the error rate of boarding station information is reduced from 6.2% to 1.0%, resulting in a 5.2%p improvement in the accuracy of smart card data. It is expected that the proposed algorithm will enhance the process of adjusting bus routes and making decisions related to public transportation infrastructure investments.

• KIPS Transactions on Computer and Communication Systems
• /
• v.2 no.8
• /
• pp.335-342
• /
• 2013

Nowadays, the traffic type and behavior are extremely diverse due to the growth of network speed and the appearance of various services on Internet. For efficient network operation and management, the importance of application-level traffic identification is more and more increasing in the area of traffic analysis. In recent years traffic identification methodology using statistical features of traffic flow has been broadly studied. However, there are several problems to be considered in the identification methodology base on statistical features of flow to improve the analysis accuracy. In this paper, we recognize these problems by analyzing the ground-truth traffic and propose the solution of these problems. The four problems considered in this paper are the distance measurement of features, the selection of the representative value of features, the abnormal behavior of TCP sessions, and the weight assignment to the feature. The proposed solutions were verified by showing the performance improvement through experiments in campus network.

• Journal of the Korean Society of Radiology
• /
• v.18 no.2
• /
• pp.93-99
• /
• 2024

The digital medical imaging, especially, computed tomography (CT), should necessarily be considered in terms of noise distribution caused by converting to X-ray photon to digital imaging signal. Recently, the denoising technique based on deep learning architecture is increasingly used in the medical imaging field. Here, we evaluated noise reduction effect according to various noise types based on the U-net deep learning model in the lung CT images. The input data for deep learning was generated by applying Gaussian noise, Poisson noise, salt and pepper noise and speckle noise from the ground truth (GT) image. In particular, two types of Gaussian noise input data were applied with standard deviation values of 30 and 50. There are applied hyper-parameters, which were Adam as optimizer function, 100 as epochs, and 0.0001 as learning rate, respectively. To analyze the quantitative values, the mean square error (MSE), the peak signal to noise ratio (PSNR) and coefficient of variation (COV) were calculated. According to the results, it was confirmed that the U-net model was effective for noise reduction all of the set conditions in this study. Especially, it showed the best performance in Gaussian noise.

• Korean Journal of Environment and Ecology
• /
• v.32 no.6
• /
• pp.686-697
• /
• 2018

Due to the advance in remote sensing technology, it has become easier to more frequently obtain high resolution imagery to detect delicate changes in an extensive area, particularly including forest which is not readily sub-classified. Time-series analysis on high resolution images requires to collect extensive amount of ground truth data. In this study, the potential of land coverage mapas ground truth data was tested in classifying high-resolution imagery. The study site was Wonju-si at Gangwon-do, South Korea, having a mix of urban and natural areas. KOMPSAT-3A imagery taken on March 2015 and land coverage map published in 2017 were used as source data. Two pixel-based classification algorithms, Support Vector Machine (SVM) and Random Forest (RF), were selected for the analysis. Forest only classification was compared with that of the whole study area except wetland. Confusion matrixes from the classification presented that overall accuracies for both the targets were higher in RF algorithm than in SVM. While the overall accuracy in the forest only analysis by RF algorithm was higher by 18.3% than SVM, in the case of the whole region analysis, the difference was relatively smaller by 5.5%. For the SVM algorithm, adding the Majority analysis process indicated a marginal improvement of about 1% than the normal SVM analysis. It was found that the RF algorithm was more effective to identify the broad-leaved forest within the forest, but for the other classes the SVM algorithm was more effective. As the two pixel-based classification algorithms were tested here, it is expected that future classification will improve the overall accuracy and the reliability by introducing a time-series analysis and an object-based algorithm. It is considered that this approach will contribute to improving a large-scale land planning by providing an effective land classification method on higher spatial and temporal scales.

• Journal of Korea Water Resources Association
• /
• v.47 no.12
• /
• pp.1227-1237
• /
• 2014

Dual-polarization can distinguish precipitation type and dual-polarization is provide not only meteorological phenomena in the atmosphere but also non-precipitation echoes. Therefore dual-polarization radar can improve radar estimates of rainfall. However polarimetric measurements by transmitting vertically vibration waves and horizontally vibrating waves simultaneously is contain systematic bias of the radar itself. Thus the calibration bias is necessary to improve quantitative precipitation estimation. In this study, the calibration bias of reflectivity (Z) and differential reflectivity ($Z_{DR}$) from the Bislsan dual-polarization radar is calculated using the 2-Dimensional Video Disdrometer (2DVD) data. And an improvement in rainfall estimation is investigated by applying derived calibration bias. A total of 33 rainfall cases occurring in Daegu from 2011 to 2012 were selected. As a results, the calibration bias of Z is about -0.3 to 5.5 dB, and $Z_{DR}$ is about -0.1 dB to 0.6 dB. In most cases, the Bislsan radar generally observes Z and $Z_{DR}$ variables lower than the simulated variables. Before and after calibration bias, compared estimated rainfall from the dual-polarization radar with AWS rain gauge in Daegu found that the mean bias has fallen by 1.69 to 1.54 mm/hr, and the RMSE has decreased by 2.54 to 1.73 mm/hr. And estimated rainfall comparing to the surface rain gauge as ground truth, rainfall estimation is improved about 7-61%.