• Journal of Cadastre & Land InformatiX
• /
• v.52 no.1
• /
• pp.105-118
• /
• 2022

The demand for information related to 3D spatial objects model in metaverse, smart cities, digital twins, autonomous vehicles, urban air mobility will be increased. 3D model construction for spatial objects is possible with various equipments such as satellite-, aerial-, ground platforms and technologies such as modeling, artificial intelligence, image matching. However, it is not easy to quickly detect and convert spatial objects that need updating. In this study, based on spatial information (features) and attributes, using matching elements such as address code, number of floors, building name, and area, the converged building DB and the detected building DB are constructed. Both to support above and to verify the suitability of object selection that needs to be updated, one system prototype was developed. When constructing the converged building DB, the convergence of spatial information and attributes was impossible or failed in some buildings, and the matching rate was low at about 80%. It is believed that this is due to omitting of attributes about many building objects, especially in the pilot test area. This system prototype will support the establishment of an efficient drone shooting plan for the rapid update of 3D spatial objects, thereby preventing duplication and unnecessary construction of spatial objects, thereby greatly contributing to object improvement and cost reduction.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.26 no.3D
• /
• pp.409-416
• /
• 2006

Real-time traveler information disseminated through Variable Message Signs (VMS) is known to have effects on driver route choice decisions. In the past, many studies have attempted to optimize the system performance using VMS message content as the primary control variable of driver route choice. This research proposes a VMS information provision optimization model which searches the best combination of VMS message contents and display sequence to minimize the total travel time on a highway network considered. The driver route choice models under VMS information provision are developed using a stated preference (SP) survey data in order to realistically capture driver response behavior. The genetic algorithm (GA) is used to find the optimal VMS information provision strategies which consists of the VMS message contents and the sequence of message display. In the process of the GA module, the system performance is measured using micro traffic simulation. The experiment results highlight the capability of the proposed model to search the optimal solution in an efficient way. The results show that the traveler information conveyed via VMS can reduce the total travel time on a highway network. They also suggest that as the frequency of VMS message update gets shorter, a smaller number of VMS message contents performs better to reduce the total travel time, all other things being equal.

• Smart Media Journal
• /
• v.12 no.11
• /
• pp.27-35
• /
• 2023

Fourth Industrial Revolution and COVID-19 pandemic have boosted the use of Government 24 app for public service complaints in the era of non-face-to-face interactions. there has been a growing influx of complaints and improvement demands from users of public apps. Furthermore, systematic management of public apps is deemed necessary. The aim of this study is to analyze the grievances of Government 24 app users, understand the current dissatisfaction among citizens, and propose potential improvements. Data were collected from the Google Play Store from May 2, 2013, to June 30, 2023, comprising a total of 6,344 records. Among these, 1,199 records with a rating of 1 and at least one 'thumbs-up' were used for topic modeling analysis. The analysis revealed seven topics: 'Issues with certificate issuance,' 'Website functionality and UI problems,' 'User ID-related issues,' 'Update problems,' 'Government employee app management issues,' 'Budget wastage concerns ((It's not worth even a single star) or (It's a waste of taxpayers' money)),' and 'Password-related problems.' Furthermore, the overall trend of these topics showed an increase until 2021, a slight decrease in 2022, but a resurgence in 2023, underscoring the urgency of updates and management. We hope that the results of this study will contribute to the development and management of public apps that satisfy citizens in the future.

• Journal of Korea Water Resources Association
• /
• v.53 no.8
• /
• pp.627-636
• /
• 2020

Debris flow disasters primarily occur in mountainous terrains far from cities. As such, they have been underestimated to cause relatively less damage compared with other natural disasters. However, owing to urbanization, several residential areas and major facilities have been built in mountainous regions, and the frequency of debris flow disasters is steadily increasing owing to the increase in rainfall with environmental and climate changes. Thus, the risk of debris flow is on the rise. However, only a few studies have explored the characteristics of flooding and reduction measures for debris flow in areas designated as steep slopes. In this regard, it is necessary to conduct research on securing independent disaster prevention technology, suitable for the environment in South Korea and reflective of the topographical characteristics thereof, and update and improve disaster prevention information. Accordingly, this study aimed to calculate the amount of debris flow, depending on disaster prevention performance targets for regions designated as steep slopes in South Korea, and develop an independent model to not only evaluate the impact of debris flow but also identify debris barriers that are optimal for mitigating damage. To validate the reliability of the two-dimensional debris flow model developed for the evaluation of debris barriers, the model's performance was compared with that of the hydraulic model. Furthermore, a 2-D debris model was constructed in consideration of the regional characteristics around the steep slopes to analyze the flow characteristics of the debris that directly reaches the damaged area. The flow characteristics of the debris delivered downstream were further analyzed, depending on the specifications (height) and installation locations of the debris barriers employed to reduce the damage. The experimental results showed that the reliability of the developed model is satisfactory; further, this study confirmed significant performance degradation of debris barriers in areas where the barriers were installed at a slope of 20° or more, which is the slope at which debris flows occur.

• Journal of Intelligence and Information Systems
• /
• v.25 no.4
• /
• pp.53-65
• /
• 2019

Object tracking is one of important steps to achieve video-based surveillance systems. Object tracking is considered as an essential task similar to object detection and recognition. In order to perform object tracking, various machine learning methods (e.g., least-squares, perceptron and support vector machine) can be applied for different designs of tracking systems. In general, generative methods (e.g., principal component analysis) were utilized due to its simplicity and effectiveness. However, the generative methods were only focused on modeling the target object. Due to this limitation, discriminative methods (e.g., binary classification) were adopted to distinguish the target object and the background. Among the machine learning methods for binary classification, total error rate minimization can be used as one of successful machine learning methods for binary classification. The total error rate minimization can achieve a global minimum due to a quadratic approximation to a step function while other methods (e.g., support vector machine) seek local minima using nonlinear functions (e.g., hinge loss function). Due to this quadratic approximation, the total error rate minimization could obtain appropriate properties in solving optimization problems for binary classification. However, this total error rate minimization was based on a batch mode setting. The batch mode setting can be limited to several applications under offline learning. Due to limited computing resources, offline learning could not handle large scale data sets. Compared to offline learning, online learning can update its solution without storing all training samples in learning process. Due to increment of large scale data sets, online learning becomes one of essential properties for various applications. Since object tracking needs to handle data samples in real time, online learning based total error rate minimization methods are necessary to efficiently address object tracking problems. Due to the need of the online learning, an online learning based total error rate minimization method was developed. However, an approximately reweighted technique was developed. Although the approximation technique is utilized, this online version of the total error rate minimization could achieve good performances in biometric applications. However, this method is assumed that the total error rate minimization can be asymptotically achieved when only the number of training samples is infinite. Although there is the assumption to achieve the total error rate minimization, the approximation issue can continuously accumulate learning errors according to increment of training samples. Due to this reason, the approximated online learning solution can then lead a wrong solution. The wrong solution can make significant errors when it is applied to surveillance systems. In this paper, we propose an exactly reweighted technique to recursively update the solution of the total error rate minimization in online learning manner. Compared to the approximately reweighted online total error rate minimization, an exactly reweighted online total error rate minimization is achieved. The proposed exact online learning method based on the total error rate minimization is then applied to object tracking problems. In our object tracking system, particle filtering is adopted. In particle filtering, our observation model is consisted of both generative and discriminative methods to leverage the advantages between generative and discriminative properties. In our experiments, our proposed object tracking system achieves promising performances on 8 public video sequences over competing object tracking systems. The paired t-test is also reported to evaluate its quality of the results. Our proposed online learning method can be extended under the deep learning architecture which can cover the shallow and deep networks. Moreover, online learning methods, that need the exact reweighting process, can use our proposed reweighting technique. In addition to object tracking, the proposed online learning method can be easily applied to object detection and recognition. Therefore, our proposed methods can contribute to online learning community and object tracking, detection and recognition communities.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.24 no.5
• /
• pp.482-488
• /
• 2014

This paper aims to analyze user's emotion automatically by analyzing Twitter, a representative social network service (SNS). In order to create sentiment analysis models by using machine learning techniques, sentiment labels that represent positive/negative emotions are required. However it is very expensive to obtain sentiment labels of tweets. So, in this paper, we propose a sentiment analysis model by using self-training technique in order to utilize "data without sentiment labels" as well as "data with sentiment labels". Self-training technique is that labels of "data without sentiment labels" is determined by utilizing "data with sentiment labels", and then updates models using together with "data with sentiment labels" and newly labeled data. This technique improves the sentiment analysis performance gradually. However, it has a problem that misclassifications of unlabeled data in an early stage affect the model updating through the whole learning process because labels of unlabeled data never changes once those are determined. Thus, labels of "data without sentiment labels" needs to be carefully determined. In this paper, in order to get high performance using self-training technique, we propose 3 policies for updating "data with sentiment labels" and conduct a comparative analysis. The first policy is to select data of which confidence is higher than a given threshold among newly labeled data. The second policy is to choose the same number of the positive and negative data in the newly labeled data in order to avoid the imbalanced class learning problem. The third policy is to choose newly labeled data less than a given maximum number in order to avoid the updates of large amount of data at a time for gradual model updates. Experiments are conducted using Stanford data set and the data set is classified into positive and negative. As a result, the learned model has a high performance than the learned models by using "data with sentiment labels" only and the self-training with a regular model update policy.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.42 no.6
• /
• pp.1-8
• /
• 2005

In this paper, we propose a two-stage head tracking algorithm adequate for real-time active camera system having pan-tilt-zoom functions. In the color convergence stage, we first assume that the shape of a head is an ellipse and its model color histogram is acquired in advance. Then, the min-shift method is applied to roughly estimate a target position by examining the histogram similarity of the model and a candidate ellipse. To reflect the temporal change of object color and enhance the reliability of mean-shift based tracking, the target histogram obtained in the previous frame is considered to update the model histogram. In the updating process, to alleviate error-accumulation due to outliers in the target ellipse of the previous frame, the target histogram in the previous frame is obtained within an ellipse adaptively shrunken on the basis of the model histogram. In addition, to enhance tracking reliability further, we set the initial position closer to the true position by compensating the global motion, which is rapidly estimated on the basis of two 1-D projection datasets. In the subsequent stage, we refine the position and size of the ellipse obtained in the first stage by using shape information. Here, we define a robust shape-similarity function based on the gradient direction. Extensive experimental results proved that the proposed algorithm performs head hacking well, even when a person moves fast, the head size changes drastically, or the background has many clusters and distracting colors. Also, the propose algorithm can perform tracking with the processing speed of about 30 fps on a standard PC.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.31 no.6_1
• /
• pp.455-461
• /
• 2013

Kompsat-3 is an optical high-resolution earth observation satellite launched in May 2012. The AEISS sensor of the Korean satellite provides 0.7m panchromatic and 2.8m multi-spectral images with 16.8km swath width from the sun-synchronous near-circular orbit of 685km altitude. Kompsat-3 is more advanced than Kompsat-2 and the improvements include better agility such as in-track stereo acquisition capability. This study investigated the characteristic of the epipolar curves of in-track Kompsat-3 stereo images. To this end we used the RPCs(Rational Polynomial Coefficients) to derive the epipolar curves over the entire image area and found out that the third order polynomial equation is required to model the curves. In addition, we could observe two different groups of curve patterns due to the dual CCDs of AEISS sensor. From the experiment we concluded that the third order polynomial-based RPCs update is required to minimize the sample direction image distortion. Finally we carried out the experiment on the epipolar resampling and the result showed the third order polynomial image transformation produced less than 0.7 pixels level of y-parallax.

• Proceedings of the Korean Society for Bioinformatics Conference
• /
• 2005.09a
• /
• pp.325-330
• /
• 2005

We are generally interested in the analysis, detection and prediction of structural motifs in proteins, in order to infer compatibility of amino acid sequence to structure in proteins of known three-dimensional structure available in the Protein Data Bank. In this context, we are analyzing some of the well-characterized structural motifs in proteins. We have analyzed simple structural motifs, such as, ${\beta}$-turns and ${\gamma}$-turns by evaluating the statistically significant type-dependent amino acid positional preferences in enlarged representative protein datasets and revised the amino acid preferences. In doing so, we identified a number of ‘unexpected’ isolated ${\beta}$-turns with a proline amino acid residue at the (i+2) position. We extended our study to the identification of multiple turns, continuous turns and to peptides that correspond to the combinations of individual ${\beta}$ and ${\gamma}$-turns in proteins and examined the hydrogen-bond interactions likely to stabilize these peptides. This led us to develop a database of structural motifs in proteins (DSMP) that would primarily allow us to make queries based on the various fields in the database for some well-characterized structural motifs, such as, helices, ${\beta}$-strands, turns, ${\beta}$-hairpins, ${\beta}$-${\alpha}$-${\beta}$, ${\psi}$-loops, ${\beta}$-sheets, disulphide bridges. We have recently implemented this information for all entries in the current PDB in a relational database called ODSMP using Oracle9i that is easy to update and maintain and added few additional structural motifs. We have also developed another relational database corresponding to amino acid sequences and their associated secondary structure for representative proteins in the PDB called PSSARD. This database allows flexible queries to be made on the compatibility of amino acid sequences in the PDB to ‘user-defined’ super-secondary structure conformation and vice-versa. Currently, we have extended this database to include nearly 23,000 protein crystal structures available in the PDB. Further, we have analyzed the ‘structural plasticity’ associated with the ${\beta}$-propeller structural motif We have developed a method to automatically detect ${\beta}$-propellers from the PDB codes. We evaluated the accuracy and consistency of predicting ${\beta}$ and ${\gamma}$-turns in proteins using the residue-coupled model. I will discuss results of our work and describe databases and software applications that have been developed.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.12 no.6
• /
• pp.629-635
• /
• 2019

In Deep Learning models derivative is implemented by error back-propagation which enables the model to learn the error and update parameters. It can find the global (or local) optimal points of parameters even in the complex models taking advantage of a huge improvement in computing power. However, deliberately generated data points can 'fool' models and degrade the performance such as prediction accuracy. Not only these adversarial examples reduce the performance but also these examples are not easily detectable with human's eyes. In this work, we propose the method to detect adversarial datasets with random noise addition. We exploit the fact that when random noise is added, prediction accuracy of non-adversarial dataset remains almost unchanged, but that of adversarial dataset changes. We set attack methods (FGSM, Saliency Map) and noise level (0-19 with max pixel value 255) as independent variables and difference of prediction accuracy when noise was added as dependent variable in a simulation experiment. We have succeeded in extracting the threshold that separates non-adversarial and adversarial dataset. We detected the adversarial dataset using this threshold.