• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.3
• /
• pp.1100-1118
• /
• 2021

In a wide range of ML applications, the training data contains privacy-sensitive information that should be kept secure. Training the ML systems by privacy-sensitive data makes the ML model inherent to the data. As the structure of the model has been fine-tuned by training data, the model can be abused for accessing the data by the estimation in a reverse process called model inversion attack (MIA). Although, MIA has been applied to shallow neural network models of recognizers in literature and its threat in privacy violation has been approved, in the case of a deep learning (DL) model, its efficiency was under question. It was due to the complexity of a DL model structure, big number of DL model parameters, the huge size of training data, big number of registered users to a DL model and thereof big number of class labels. This research work first analyses the possibility of MIA on a deep learning model of a recognition system, namely a face recognizer. Second, despite the conventional MIA under the white box scenario of having partial access to the users' non-sensitive information in addition to the model structure, the MIA is implemented on a deep face recognition system by just having the model structure and parameters but not any user information. In this aspect, it is under a semi-white box scenario or in other words a gray-box scenario. The experimental results in targeting five registered users of a CNN-based face recognition system approve the possibility of regeneration of users' face images even for a deep model by MIA under a gray box scenario. Although, for some images the evaluation recognition score is low and the generated images are not easily recognizable, but for some other images the score is high and facial features of the targeted identities are observable. The objective and subjective evaluations demonstrate that privacy cyber-attack by MIA on a deep recognition system not only is feasible but also is a serious threat with increasing alert state in the future as there is considerable potential for integration more advanced ML techniques to MIA.

• Korean Journal of Remote Sensing
• /
• v.39 no.1
• /
• pp.77-86
• /
• 2023

Ship detection is widely used in areas such as maritime security, maritime traffic, fisheries management, illegal fishing, and border control, and ship detection is important for rapid response and damage minimization as ship accident rates increase due to recent increases in international maritime traffic. Currently, according to a number of global and national regulations, ships must be equipped with automatic identification system (AIS), which provide information such as the location and speed of the ship periodically at regular intervals. However, most small vessels (less than 300 tons) are not obligated to install the transponder and may not be transmitted intentionally or accidentally. There is even a case of misuse of the ship'slocation information. Therefore, in this study, ship detection was performed using high-resolution optical satellite images that can periodically remotely detect a wide range and detectsmallships. However, optical images can cause false-alarm due to noise on the surface of the sea, such as waves, or factors indicating ship-like brightness, such as clouds and wakes. So, it is important to remove these factors to improve the accuracy of ship detection. In this study, false alarm wasreduced, and the accuracy ofship detection wasimproved by removing wake.As a ship detection method, ship detection was performed using machine learning-based random forest (RF), and convolutional neural network (CNN) techniquesthat have been widely used in object detection fieldsrecently, and ship detection results by the model were compared and analyzed. In addition, in this study, the results of RF and CNN were combined to improve the phenomenon of ship disconnection and the phenomenon of small detection. The ship detection results of thisstudy are significant in that they improved the limitations of each model while maintaining accuracy. In addition, if satellite images with improved spatial resolution are utilized in the future, it is expected that ship and wake simultaneous detection with higher accuracy will be performed.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.45 no.2
• /
• pp.48-55
• /
• 2022

The color image of the brand comes first and is an important visual element that leads consumers to the consumption of the product. To express more effectively what the brand wants to convey through design, the printing market is striving to print accurate colors that match the intention. In 'offset printing' mainly used in printing, colors are often printed in CMYK (Cyan, Magenta, Yellow, Key) colors. However, it is possible to print more accurate colors by making ink of the desired color instead of dotting CMYK colors. The resulting ink is called 'spot color' ink. Spot color ink is manufactured by repeating the process of mixing the existing inks. In this repetition of trial and error, the manufacturing cost of ink increases, resulting in economic loss, and environmental pollution is caused by wasted inks. In this study, a deep learning algorithm to predict printed spot colors was designed to solve this problem. The algorithm uses a single DNN (Deep Neural Network) model to predict printed spot colors based on the information of the paper and the proportions of inks to mix. More than 8,000 spot color ink data were used for learning, and all color was quantified by dividing the visible light wavelength range into 31 sections and the reflectance for each section. The proposed algorithm predicted more than 80% of spot color inks as very similar colors. The average value of the calculated difference between the actual color and the predicted color through 'Delta E' provided by CIE is 5.29. It is known that when Delta E is less than 10, it is difficult to distinguish the difference in printed color with the naked eye. The algorithm of this study has a more accurate prediction ability than previous studies, and it can be added flexibly even when new inks are added. This can be usefully used in real industrial sites, and it will reduce the attempts of the operator by checking the color of ink in a virtual environment. This will reduce the manufacturing cost of spot color inks and lead to improved working conditions for workers. In addition, it is expected to contribute to solving the environmental pollution problem by reducing unnecessarily wasted ink.

• Journal of the Korean Society for information Management
• /
• v.40 no.3
• /
• pp.143-162
• /
• 2023

This study analyzed the Beopmaru, Supreme Court Library of Korea, circulation data to identify user lending patterns and proposed a plan to reflect the analysis results in future user services. In 2022, Beopmaru's collection of books was 212,608, with law books accounting for 73%. However, general books accounted for 83% of actual circulation. Looking at the usage coefficient by topic, the literature field was the most actively used at 5.85, and the law field was the least used at 0.23. In the case of interlibrary loan, both KERIS member institutions and the Korean Bar Association had the highest loan ratios in the legal field, civil law field, and judicial litigation procedure field, in that order. However, member institutions affiliated with KERIS, a legal academic community, were lending law books on a wider range of subject areas than the Korean Bar Association, a practical organization. To improve access to legal information, the Beopmaru public service was implemented, but in reality, the use of reading space was high, and the proportion of general books loaned was much higher. In order to improve this, it seems necessary to strengthen the promotion of Beopmaru loan services, provide personalized services, improve book lending regulations, strengthen online services, and establish a cooperative network.

• Journal of the Korea Society for Simulation
• /
• v.19 no.4
• /
• pp.77-87
• /
• 2010

Currently, our society has been changed from the industrial society to the information society. As the war progresses to Information Warfare, Network-Centric Warfare, Long-Range Precision Engagement and Robot Warfare, the military should advance to High-tech Scientific force. For this creation of the war potential, it is regarded as the warfighting experiment is a critical method. Surely it is rational that LVC(Live Virtual Constructive simulation) is desirable to make the warfighting experiment. But because it is limited by the cost, the time, the place and the resource, the constructive simulation(M&S : Modeling&Simulation) is a good tool to solve those problems. There are some studies about the evaluation process for developing the model, but it is unsatisfying in the process of the constructive simulations' operation. This study focuses on the way of constructive simulation operation, which is supplied with the evaluation process(VV&A : Verification Validation & Accreditation). We introduce the example of the rear area operation simulation for "appropriateness evaluation to the organization of logistic corps" by the AWAM(Army Weapon Analysis Model). This study presents the effective methods of the constructive simulations, which is based on the reliable evaluation process, so it will contribute to the warfighting experiments.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.8
• /
• pp.103-109
• /
• 2023

In this paper, we propose a smart wall switch based on IEEE 802.11b/g/n standard 2.4GHz band communication. As the 4th industrial era evolves, smart home solution development is actively underway, and application cases for smart wall switches are increasing. Most of the Chinese products that preoccupy the market through price competitiveness use Bluetooth and Zigbee communication switches. However, while ZigBee communication is low power, communication speed is slower than Bluetooth and network configuration through a separate hub is additionally required. The Bluetooth method has problems in that the communication range and speed are lower than Wi-Fi communication, the communication standby time is relatively long, and security is weak. In this study, an IEEE 802.11b/g/n smart wall switch applied with Wi-Fi communication technology was developed. In addition, through the two-wire structure, it is designed so that no additional cost is incurred through the construction of a separate neutral line in the building. The result of the study is more than 30% cheaper than the existing wall switch, so it is judged that it will be able to preoccupy the market not only in terms of technological competitiveness but also price competitiveness.

• Journal of Platform Technology
• /
• v.11 no.4
• /
• pp.19-34
• /
• 2023

In order to identify the trends and insights of 'startups' in the global era, we conducted an in-depth trend analysis of the global startup ecosystem using Google Trends, a big data analysis platform. For the validity of the analysis, we verified the correlation between the keywords 'startup' and 'global' through BIGKinds. We also conducted a network analysis based on the data extracted using Google Trends to determine the frequency of searches for the keyword or term 'startup'. The results showed a strong positive linear relationship between the keywords, indicating a statistically significant correlation (correlation coefficient: +0.8906). When exploring global startup trends using Google Trends, we found a terribly similar linear pattern of increasing and decreasing interest in each country over time, as shown in Figure 4. In particular, startup interest was low in the range of 35 to 76 from mid-2020 due to the COVID-19 pandemic, but there was a noticeable upward trend in startup interest after March 2022. In addition, we found that the interest in startups in each country except South Korea is very similar, and the related topics are startup company, technology, investment, funding, and keyword search terms such as best startup, tech, business, invest, health, and fintech are highly correlated.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.28 no.5A
• /
• pp.673-683
• /
• 2008

This study improved the existing method of using the longitudinal strain and concept of influence line to develop Bridge Weigh-in-Motion system without axle detector using the dynamic strain of the bridge girders and concrete slab. This paper first describes the considered algorithms of extracting passing vehicle information from the dynamic strain signal measured at the bridge slab, girders, and cross beams. Two different analysis methods of 1) influence line method, and 2) neural network method are considered, and parameter study of measurement locations is also performed. Then the procedures and the results of field tests are described. The field tests are performed to acquire training sets and test sets for neural networks, and also to verify and compare performances of the considered algorithms. Finally, comparison between the results of different algorithms and discussions are followed. For a PSC I-girder bridge, vehicle weight can be calculated within a reasonable error range using the dynamic strain gauge installed on the girders. The passing lane and passing speed of the vehicle can be accurately estimated using the strain signal from the concrete slab. The passing speed and peak duration were added to the input variables to reflect the influence of the dynamic interaction between the bridge and vehicles, and impact of the distance between axles, respectively; thus improving the accuracy of the weight calculation.

• Smart Structures and Systems
• /
• v.30 no.5
• /
• pp.521-535
• /
• 2022

Efficient management of deteriorating civil infrastructure is one of the most important research topics in many developed countries. In particular, the remote displacement measurement of bridges using linear variable differential transformers, global positioning systems, laser Doppler vibrometers, and computer vision technologies has been attempted extensively. This paper proposes a remote displacement measurement system using closed-circuit televisions (CCTVs) and a computer-vision-based method for in-situ bridge bearings having relatively large displacement due to temperature change in long term. The hardware of the system is composed of a reference target for displacement measurement, a CCTV to capture target images, a gateway to transmit images via a mobile network, and a central server to store and process transmitted images. The usage of CCTV capable of night vision capture and wireless data communication enable long-term 24-hour monitoring on wide range of bridge area. The computer vision algorithm to estimate displacement from the images involves image preprocessing for enhancing the circular features of the target, circular Hough transformation for detecting circles on the target in the whole field-of-view (FOV), and homography transformation for converting the movement of the target in the images into an actual expansion displacement. The simple target design and robust circle detection algorithm help to measure displacement using target images where the targets are far apart from each other. The proposed system is installed at the Tancheon Overpass located in Seoul, and field experiments are performed to evaluate the accuracy of circle detection and displacement measurements. The circle detection accuracy is evaluated using 28,542 images captured from 71 CCTVs installed at the testbed, and only 48 images (0.168%) fail to detect the circles on the target because of subpar imaging conditions. The accuracy of displacement measurement is evaluated using images captured for 17 days from three CCTVs; the average and root-mean-square errors are 0.10 and 0.131 mm, respectively, compared with a similar displacement measurement. The long-term operation of the system, as evaluated using 8-month data, shows high accuracy and stability of the proposed system.

• Journal of Internet Computing and Services
• /
• v.24 no.6
• /
• pp.91-98
• /
• 2023

Recently, according to development of artificial intelligence, a wide range of industry being automatic and optimized. Also we can find out some research of using supervised learning for deteceting defect of railway in domestic rail industry. However, there are structures other than rails on the track, and the fastener is a device that binds the rail to other structures, and periodic inspections are required to prevent safety accidents. In this paper, we present a method of reducing cost for labeling using semi-supervised and transfer model trained on rail fastener data. We use Resnet50 as the backbone network pretrained on ImageNet. At first we randomly take training data from unlabeled data and then labeled that data to train model. After predict unlabeled data by trained model, we adopted a method of adding the data with the highest probability for each class to the training data by a predetermined size. Futhermore, we also conducted some experiments to investigate the influence of the number of initially labeled data. As a result of the experiment, model reaches 92% accuracy which has a performance difference of around 5% compared to supervised learning. This is expected to improve the performance of the classifier by using relatively few labels without additional labeling processes through the proposed method.