• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.7 no.3
• /
• pp.329-337
• /
• 2017

This paper proposes a search method that explores parameters C and σ values of support vector machines (SVM) to improve performance while maintaining search accuracy. A traditional grid search method requires tremendous computational times because it searches all available combinations of C and σ values to find optimal combinations which provide the best performance of SVM. To address this issue, this paper proposes a deep search method that reduces computational time. In the first stage, it divides C-σ- accurate metrics into four regions, searches a median value of each region, and then selects a point of the highest accurate value as a start point. In the second stage, the selected start points are re-divided into four regions, and then the highest accurate point is assigned as a new search point. In the third stage, after eight points near the search point. are explored and the highest accurate value is assigned as a new search point, corresponding points are divided into four parts and it calculates an accurate value. In the last stage, it is continued until an accurate metric value is the highest compared to the neighborhood point values. If it is not satisfied, it is repeated from the second stage with the input level value. Experimental results using normal and defect bearings show that the proposed deep search algorithm outperforms the conventional algorithms in terms of performance and search time.

• Tunnel and Underground Space
• /
• v.33 no.3
• /
• pp.109-125
• /
• 2023

Disposal repository for high-level radioactive waste secures its safety by means of engineered and natural barriers. The performance of these barriers should be tested and verified through various aspects in terms of short and/or long-term. KAERI has been conducting various in-situ demonstrations in KURT (KAERI Underground Research Tunnel). After completing previous experiment, a conceptual design of an improved in-situ experiment, i.e. K-COIN (KURT experiment of THMC COupled and INteraction), was established and detailed planning for the experiment is underway. Preliminary characterizations were conducted in KURT for siting a K-COIN test site. 15 boreholes with a depth of about 20 m were drilled in three research galleries in KURT and intact rock specimens were prepared for laboratory tests. Using the specimens, physical measurements, uniaxial compression, indirect tension, and triaxial compression tests were conducted. As a result, specific gravity, porosity, elastic wave velocities, uniaxial compressive strength, Young's modulus, Poisson's ratio, Brazilian tensile strength, cohesion, and internal friction angle were estimated. Statistical analyses revealed that there did not exist meaningful differences in intact rock properties according to the drilled sites and the depth. Judging from the uniaxial compressive strength, which is one of the most important properties, all the specimens were classified as very strong rock so that mechanical safety was secured in all the regions.

• Journal of the Korea Institute of Building Construction
• /
• v.23 no.4
• /
• pp.465-476
• /
• 2023

This study introduces the development of a precast concrete double-wall, applicable to basement construction in apartment buildings. Unlike traditional precast concrete double walls, the developed double-wall doesn't require specialized manufacturing equipment such as a lathe. The constructability of these advanced technologies was validated through a full-scale mock-up test using the precast concrete double wall. The test specimens were constructed to represent a structural wall with a thickness of 250mm. It was observed that the quality of the in-situ concrete, filled between two single panels of 110mm thickness each, was excellent. The construction efficiency of the developed double-wall system for basement construction in an apartment building was also examined. Expert interviews about installation times of precast concrete elements were conducted to evaluate the speed of the basement floor's installation. The results showed that installation of precast concrete elements, including the proposed double walls, could be completed within 20 to 29 days for a basement in an apartment building. This indicates a three-fold increase in construction efficiency compared to traditional methods relying on in-situ casting.

• Journal of the Korea Computer Graphics Society
• /
• v.29 no.3
• /
• pp.85-92
• /
• 2023

Recently, interest in artificial intelligence has increased due to the development of artificial intelligence fields such as ChatGPT and self-driving cars. However, there are still many unknown elements in training process of artificial intelligence, so that optimizing the model requires more time and effort than it needs. Therefore, there is a need for a tool or methodology that can analyze the weight changes during the training process of artificial intelligence and help out understatnding those changes. In this research, I propose a visualization system which helps people to understand the accumulated weight changes. The system calculates the weights for each training period to accumulates weight changes and stores accumulated weight changes to plot them in 3D space. This research will allow us to explore different aspect of artificial intelligence learning process, such as understanding how the model get trained and providing us an indicator on which hyperparameters should be changed for better performance. These attempts are expected to explore better in artificial intelligence learning process that is still considered as unknown and contribute to the development and application of artificial intelligence models.

• Convergence Security Journal
• /
• v.22 no.2
• /
• pp.131-141
• /
• 2022

Future battlefield domains are expanding to ground, sea, air, space, and cyber, so future military operations are expected to be carried out simultaneously and complexly in various battlefield domains. In addition, the application of convergence technologies that create innovations in all fields of economy, society, and defense, such as artificial intelligence, IoT, and big data, is being promoted. However, since the current Korean military C4I system manages warfighting function DBs in one DB server, the efficiency of combat performance is reduced utilization and distribution speed of data and operation response time. To solve this problem, research is needed on how to apply the 4th industrial revolution technologies such as AI, IoT, 5G, big data, and cloud to the Korean military C4I system, but research on this is insufficient. Therefore, this paper analyzes the problems of the current Korean military C4I system and proposes to apply the 4th industrial revolution technology in terms of operational mission, network and data link, computing environment, cyber operation, interoperability and interlocking capabilities.

• The Journal of Bigdata
• /
• v.7 no.2
• /
• pp.31-41
• /
• 2022

Steel is one of the most fundamental components to mechanical industry. However, the quality of products are greatly impacted by the surface defects in the steel. Thus, researchers pay attention to the need for surface defects detector and the deep learning methods are the current trend of object detector. There are still limitations and rooms for improvements, for example, related works focus on developing the models but don't take into account real-time application with practical implication on industrial settings. In this paper, a real-time application of steel surface defects detection based on YOLOv4 is proposed. Firstly, as the aim of this work to deploying model on real-time application, we studied related works on this field, particularly focusing on one-stage detector and YOLO algorithm, which is one of the most famous algorithm for real-time object detectors. Secondly, using pre-trained Yolov4-Darknet platform models and transfer learning, we trained and test on the hot rolled steel defects open-source dataset NEU-DET. In our study, we applied our application with 4 types of typical defects of a steel surface, namely patches, pitted surface, inclusion and scratches. Thirdly, we evaluated YOLOv4 trained model real-time performance to deploying our system with accuracy of 87.1 % mAP@0.5 and over 60 fps with GPU processing.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.4
• /
• pp.21-29
• /
• 2023

In this paper, we propose an AI-based system for automatically classifying industry and occupation codes in the population census. The accurate classification of industry and occupation codes is crucial for informing policy decisions, allocating resources, and conducting research. However, this task has traditionally been performed by human coders, which is time-consuming, resource-intensive, and prone to errors. Our system represents a significant improvement over the existing rule-based system used by the statistics agency, which relies on user-entered data for code classification. In this paper, we trained and evaluated several models, and developed an ensemble model that achieved an 86.76% match accuracy in industry and 81.84% in occupation, outperforming the best individual model. Additionally, we propose process improvement work based on the classification probability results of the model. Our proposed method utilizes an ensemble model that combines transfer learning techniques with pre-trained models. In this paper, we demonstrate the potential for AI-based systems to improve the accuracy and efficiency of population census data classification. By automating this process with AI, we can achieve more accurate and consistent results while reducing the workload on agency staff.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.21 no.1
• /
• pp.240-257
• /
• 2022

Autonomous vehicles use various local sensors such as camera, radar, and lidar to perceive the surrounding environment. However, it is difficult to predict the movement of vulnerable road users using only local sensors that are subject to limits in cognitive range. This is true especially when these users are blocked from view by obstacles. Hence, this paper developed an algorithm for collision avoidance with VRU using V2X information. The main purpose of this collision avoidance system is to overcome the limitations of the local sensors. The algorithm first evaluates the risk of collision, based on the current driving condition and the V2X information of the VRU. Subsequently, the algorithm takes one of four evasive actions; steering, braking, steering after braking, and braking after steering. A simulation was performed under various conditions. The results of the simulation confirmed that the algorithm could significantly improve the performance of the collision avoidance system while securing vehicle stability during evasive maneuvers.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.7
• /
• pp.303-314
• /
• 2023

Voice conversion, a technology that allows an individual's speech data to be regenerated with the acoustic properties(tone, cadence, gender) of another, has countless applications in education, communication, and entertainment. This paper proposes an approach based on the StarGAN-VC model that generates realistic-sounding speech without requiring parallel utterances. To overcome the constraints of the existing StarGAN-VC model that utilizes one-hot vectors of original and target speaker information, this paper extracts feature vectors of target speakers using a pre-trained version of Rawnet3. This results in a latent space where voice conversion can be performed without direct speaker-to-speaker mappings, enabling an any-to-any structure. In addition to the loss terms used in the original StarGAN-VC model, Wasserstein distance is used as a loss term to ensure that generated voice segments match the acoustic properties of the target voice. Two Time-Scale Update Rule (TTUR) is also used to facilitate stable training. Experimental results show that the proposed method outperforms previous methods, including the StarGAN-VC network on which it was based.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.30 no.2A
• /
• pp.93-102
• /
• 2010

The numerical analysis and safety assessment by construction stages were considered the essential examination particular in order to solving the unstability of long-span bridges in the middle a construction. When estimating structural response characteristics by the construction stage analysis of long-span bridges, the influence of the near-field ground motion (NFGM) would be evaluated as a critical factor for the seismic design because it indicates clearly different aspects from the existing input earthquake motion data. Therefore, this study re-examined the response aspect of long-span bridges considering NFGM characteristics based on the response spectrum result, and advanced the presented numerical analysis program by the related research for conducting the construction stage analysis and reliability assessment of long-span bridges efficiently. The excellency of various construction schemes was assessed using the time history analysis result of critical member considering NFGM characteristics. For evaluating quantitative safety level, the reliability analysis was conducted considering the influence of external uncertainties included in random variables, and presented the safety index and failure probability of the critical construction stage by NFGM characteristics. In addition, the reliability result was examined the influence of internal uncertainties using monte carlo simulation (MCS), and assessed the distribution aspect of the essential analysis result. It is expected that this study will provide the basic information for the construction safety improvement when performing seismic design of long-span bridges considering NFGM characteristics.