• Journal of Intelligence and Information Systems
• /
• v.28 no.2
• /
• pp.101-125
• /
• 2022

Recently, with the development of computing technology and the improvement of the cloud environment, deep learning technology has developed, and attempts to apply deep learning to various fields are increasing. A typical example is anomaly detection, which is a technique for identifying values or patterns that deviate from normal data. Among the representative types of anomaly detection, it is very difficult to detect a contextual anomaly that requires understanding of the overall situation. In general, detection of anomalies in image data is performed using a pre-trained model trained on large data. However, since this pre-trained model was created by focusing on object classification of images, there is a limit to be applied to anomaly detection that needs to understand complex situations created by various objects. Therefore, in this study, we newly propose a two-step pre-trained model for detecting abnormal situation. Our methodology performs additional learning from image captioning to understand not only mere objects but also the complicated situation created by them. Specifically, the proposed methodology transfers knowledge of the pre-trained model that has learned object classification with ImageNet data to the image captioning model, and uses the caption that describes the situation represented by the image. Afterwards, the weight obtained by learning the situational characteristics through images and captions is extracted and fine-tuning is performed to generate an anomaly detection model. To evaluate the performance of the proposed methodology, an anomaly detection experiment was performed on 400 situational images and the experimental results showed that the proposed methodology was superior in terms of anomaly detection accuracy and F1-score compared to the existing traditional pre-trained model.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_2
• /
• pp.1633-1641
• /
• 2022

Urban population concentration and indiscriminate development are causing various environmental problems such as air pollution and heat island phenomena, and causing human resources to deteriorate the damage caused by natural disasters. Urban trees have been proposed as a solution to these urban problems, and actually play an important role, such as providing environmental improvement functions. Accordingly, quantitative measurement and analysis of individual trees in urban trees are required to understand the effect of trees on the urban environment. However, the complexity and diversity of urban trees have a problem of lowering the accuracy of single tree detection. Therefore, we conducted a study to effectively detect trees in Dongjak-gu using high-resolution aerial images that enable effective detection of tree objects and You Only Look Once Version 5 (YOLOv5), which showed excellent performance in object detection. Labeling guidelines for the construction of tree AI learning datasets were generated, and box annotation was performed on Dongjak-gu trees based on this. We tested various scale YOLOv5 models from the constructed dataset and adopted the optimal model to perform more efficient urban tree detection, resulting in significant results of mean Average Precision (mAP) 0.663.

• Journal of Internet Computing and Services
• /
• v.22 no.6
• /
• pp.33-40
• /
• 2021

Automatic Speech Recognition(ASR) is a technology that analyzes human speech sound into speech signals and then automatically converts them into character strings that can be understandable by human. Speech recognition technology has evolved from the basic level of recognizing a single word to the advanced level of recognizing sentences consisting of multiple words. In real-time voice conversation, the high recognition rate improves the convenience of natural information delivery and expands the scope of voice-based applications. On the other hand, with the active application of speech recognition technology, concerns about related cyber attacks and threats are also increasing. According to the existing studies, researches on the technology development itself, such as the design of the Automatic Speaker Verification(ASV) technique and improvement of accuracy, are being actively conducted. However, there are not many analysis studies of attacks and threats in depth and variety. In this study, we propose a cyber attack model that bypasses voice authentication by simply manipulating voice frequency and voice speed for AI voice recognition service equipped with automated identification technology and analyze cyber threats by conducting extensive experiments on the automated identification system of commercial smartphones. Through this, we intend to inform the seriousness of the related cyber threats and raise interests in research on effective countermeasures.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.39 no.6
• /
• pp.483-490
• /
• 2021

Due to the recent increase in various problems from underground development in urbanized areas, accurate underground facility information management is highly needed. Therefore, in this study, in order to utilize the Integration Map of Underground Goespatial Information in real time on-site, the function of comparing the mutual location of the GPR (Ground Penetration Radar) sensing data and the Integration Map of Underground Goespatial Information, and function of analyze underground facilities, and function of converting surveying data into a shape file through position correction & attribute editing in a 3D space, and the function of submitting the shape file to the Integration Map of Underground Goespatial Information mobile center was defined and developed as a program. In addition, for the on-site application test of the development program, scenarios used at the underground facility real-time survey site and GPR exploration site were derived, and four sites in Seoul were tested to confirm that the use scenario worked properly. Through this, the on-site utilization of the program developed in this study could be confirmed, and it would contribute to the confirmation of the quality of Shape-file and the "update automation" of "Integration Map of Underground Goespatial Information". In addition, it is expected that the development program will be further applied to the Underground Facility Map's Accuracy Improvement Diffusion Project' promoted by the MOLIT (Ministry of Land, Infrastructure, and Transport).

• Korean Journal of Remote Sensing
• /
• v.38 no.5_3
• /
• pp.913-924
• /
• 2022

In this paper, we presented a database construction of undersea images for the Habitats of Ecklonia cava and Sargassum and conducted an experiment for semantic segmentation using state-of-the-art (SOTA) models such as High Resolution Network-Object Contextual Representation (HRNet-OCR) and Shifted Windows-L (Swin-L). The result showed that our segmentation models were superior to the existing experiments in terms of the 29% increased mean intersection over union (mIOU). Swin-L model produced better performance for every class. In particular, the information of the Ecklonia cava class that had small data were also appropriately extracted by Swin-L model. Target objects and the backgrounds were well distinguished owing to the Transformer backbone better than the legacy models. A bigger database under construction will ensure more accuracy improvement and can be utilized as deep learning database for undersea images.

• Journal of Korea Water Resources Association
• /
• v.55 no.9
• /
• pp.679-686
• /
• 2022

As the occurrences of flash floods have increased due to climate change, faster and more accurate precipitation observation using X-band radar has become important. Therefore, the Ministry of Environment installed two dual-pol X-band radars at Samcheok and Uljin. The radar data used in this study were obtained from two different elevation angles and composed to reduce the shielding effect. To obtain quantitative rainfall, quality control (QC), K_DP retrieval, and Hybrid Surface Rainfall (HSR) methods were sequentially applied. To improve the accuracy of the quantitative precipitation estimation (QPE) of the X-band radar, we retrieved parameters for the relationship between rainfall rate and specific differential phase, which is commonly called the R-K_DP relationship; hence, an empirical approach was developed using multiple rain gauges for those two radars. The newly suggested relationship, R = 27.4K^0.81_DP, slightly increased the correlation coefficient by 1% more than the relationship suggested by the previous study. The root mean square error significantly decreased from 3.88 mm/hr to 3.68 mm/hr, and the bias of the estimated precipitation also decreased from -1.72 mm/hr to -0.92 mm/hr for overall cases, showing the improvement of the new method.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.4
• /
• pp.11-19
• /
• 2021

An abnormal object refers to a person, an object, or a mechanical device that performs abnormal and unusual behavior and needs observation or supervision. In order to detect this through artificial intelligence algorithm without continuous human intervention, a method of observing the specificity of temporal features using optical flow technique is widely used. In this study, an abnormal situation is identified by learning an algorithm that translates an input image frame to an optical flow image using a Generative Adversarial Network (GAN). In particular, we propose a technique that improves the pre-processing process to exclude unnecessary outliers and the post-processing process to increase the accuracy of identification in the test dataset after learning to improve the performance of the model's abnormal behavior identification. UCSD Pedestrian and UMN Unusual Crowd Activity were used as training datasets to detect abnormal behavior. For the proposed method, the frame-level AUC 0.9450 and EER 0.1317 were shown in the UCSD Ped2 dataset, which shows performance improvement compared to the models in the previous studies.

• Korean Association of Arts Management
• /
• no.53
• /
• pp.101-125
• /
• 2020

This paper presents the problems and improvement methods in estimating demand and benefit, which have been controversial in the feasibility study of building cultural facilities. Although there are justifications for supplying cultural facilities by expanding leisure time and increasing income, the economic burden from the insolvent operation after construction is high. Feasibility studies can prevent these problems in advance. In order to estimate the demand for cultural facilities, similar facilities were selected and the gravity model was used to estimate the demand. In the future, it is necessary to prepare the criteria for setting the reference facility to increase the accuracy of the demand estimation. In addition, in the case of cultural facilities constructed through feasibility study, it is necessary to induce and enforce the disclosure of operational data and information, and to establish a database so that it can be used as a reference facility for demand estimation in future feasibility study on cultural facility. Accurate benefit estimation requires multiple CVM surveys. In addition to the current CVM survey, this paper suggest that supplementary online non-face-to-face surveys is considered. Furthermore, this research suggests that the use of video media for explanation of alternative materials for cultural facilities to be constructed because the WTP may be excessive due to lack of alternatives for survey respondents in the current CVM survey.

• Journal of the Korea Institute of Building Construction
• /
• v.22 no.6
• /
• pp.689-700
• /
• 2022

Layout work for building structures requires high precision to construct structural elements in the correct location. However, the accuracy and precision of the layout position are affected by the worker's skill, and productivity can be reduced when there is information loss and error. To solve this problem, it is necessary to automate the overall layout operation and introduce information technology, and layout process automation using construction robots can be an effective means of doing this. This study develops a prototype of an automated layout robot for building structures and evaluates its basic performance. The developed robot is largely composed of driving, marking, sensing, and control units, and is designed to enable various driving methods, and movement and rotation of the marking unit in consideration of the environment on structural work. The driving and marking performance experiments showed satisfactory performance in terms of driving distance error and marking quality, while the need for improvement in terms of some driving methods and marking precision was confirmed. Based on the results of this study, we intend to continuously improve the robot's performance and establish an automation system for overall layout work process.

• Journal of Aerospace System Engineering
• /
• v.16 no.4
• /
• pp.60-67
• /
• 2022

An Electro-Optical Tracking System (EOTS) is an electric optical system with EO/IR cameras, laser sensors, and an IMU. The EOTS calculates coordinates of targets, using attitude and acceleration measured by the IMU. In particular for an armed aircraft, the performance of the weapon system depends on how quickly and accurately it acquires the target coordinates. The IMU should be operated after alignment is complete, to meet the coordinate accuracy required by the weapon system so the initial stabilization time of the IMU should be reduced, by quickly measuring the attitude and acceleration. Alignment is the process of determining the initial attitude by resolving the attitude error of the IMU, and the IMU of mission equipment such as an airborne EOTS, uses velocity matching based on the velocity from GPS/INS for aircraft navigation. In this paper, a method is presented to improve the transfer alignment performance of the airborne EOTS, by maneuvering aircraft and the mission equipment. First, the performance factor of the alignment was identified, as a heading error through the velocity matching model and simulation results. Then acceleration maneuvers and attitude changes were necessary, to correct the error. As a result of flight tests applied to an EOTS on a OOO aircraft system, the transfer alignment performance was improved as the duration time was decreased, by more than five times when the aircraft accelerated by more than 0.2g and the EOTS was moving until 6.7deg/s.