• Journal of the Korea Institute of Military Science and Technology
• /
• v.26 no.1
• /
• pp.44-53
• /
• 2023

Adversarial attacks have received great attentions for their capacity to distract state-of-the-art neural networks by modifying objects in physical domain. Patch-based attack especially have got much attention for its optimization effectiveness and feasible adaptation to any objects to attack neural network-based object detectors. However, despite their strong attack performance, generated patches are strongly perceptible for humans, violating the fundamental assumption of adversarial examples. In this paper, we propose a camouflaged adversarial patch optimization method using military camouflage assessment metrics for naturalistic patch attacks. We also investigate camouflaged attack loss functions, applications of various camouflaged patches on army tank images, and validate the proposed approach with extensive experiments attacking Yolov5 detection model. Our methods produce more natural and realistic looking camouflaged patches while achieving competitive performance.

• Journal of Korean Society for Quality Management
• /
• v.51 no.1
• /
• pp.119-129
• /
• 2023

Purpose: This study proposes a classification model for implementing condition-based maintenance (CBM) by monitoring the real-time status of a machine using acceleration sensor data collected from a vehicle. Methods: The classification model's performance was improved by applying Fourier transform to convert the acceleration sensor data from the time domain to the frequency domain. Additionally, the Generative Adversarial Network (GAN) algorithm was used to augment images and further enhance the classification model's performance. Results: Experimental results demonstrate that the GAN algorithm can effectively serve as an image augmentation technique to enhance the performance of the classification model. Consequently, the proposed approach yielded a significant improvement in the classification model's accuracy. Conclusion: While this study focused on the effectiveness of the GAN algorithm as an image augmentation method, further research is necessary to compare its performance with other image augmentation techniques. Additionally, it is essential to consider the potential for performance degradation due to class imbalance and conduct follow-up studies to address this issue.

• Steel and Composite Structures
• /
• v.45 no.2
• /
• pp.281-291
• /
• 2022

Surmounting complexities in analyzing the mechanical parameters of concrete entails selecting an appropriate methodology. This study integrates a novel metaheuristic technique, namely satin bowerbird optimizer (SBO) with artificial neural network (ANN) for predicting uniaxial compressive strength (UCS) of concrete. For this purpose, the created hybrid is trained and tested using a relatively large dataset collected from the published literature. Three other new algorithms, namely Henry gas solubility optimization (HGSO), sunflower optimization (SFO), and vortex search algorithm (VSA) are also used as benchmarks. After attaining a proper population size for all algorithms, the Utilizing various accuracy indicators, it was shown that the proposed ANN-SBO not only can excellently analyze the UCS behavior, but also outperforms all three benchmark hybrids (i.e., ANN-HGSO, ANN-SFO, and ANN-VSA). In the prediction phase, the correlation indices of 0.87394, 0.87936, 0.95329, and 0.95663, as well as mean absolute percentage errors of 15.9719, 15.3845, 9.4970, and 8.0629%, calculated for the ANN-HGSO, ANN-SFO, ANN-VSA, and ANN-SBO, respectively, manifested the best prediction performance for the proposed model. Also, the ANN-VSA achieved reliable results as well. In short, the ANN-SBO can be used by engineers as an efficient non-destructive method for predicting the UCS of concrete.

• Journal of the Semiconductor & Display Technology
• /
• v.23 no.2
• /
• pp.107-113
• /
• 2024

Scheduling of semiconductor test facilities has been crucial since effective scheduling contributes to the profits of semiconductor enterprises and enhances the quality of semiconductor products. This study aims to solve the scheduling problems for the wafer burn-in test facilities of the semiconductor back-end process by utilizing simulation and deep reinforcement learning-based methods. To solve the scheduling problem considered in this study. we propose novel state, action, and reward designs based on the Markov decision process. Furthermore, a neural network is trained by employing the recent RL-based method, named proximal policy optimization. Experimental results showed that the proposed method outperformed traditional heuristic-based scheduling techniques, achieving a higher due date compliance rate of jobs in terms of total job completion time.

• Journal of Korean Society of Transportation
• /
• v.22 no.4 s.75
• /
• pp.135-146
• /
• 2004

For improving the effectiveness of travel information, some rational paths are needed to provide them to users driving in real road network. To meet it, k-shortest path algorithms have been used in general. Although the k-shortest path algorithm can provide several alternative paths, it has inherent limit of heavy overlapping among derived paths, which nay lead to incorrect travel information to the users. In case of considering the network consisting of several turn prohibitions popularly adopted in real world network, it makes difficult for the traditional network optimization technique to deal with. Banned and penalized turns are not described appropriately for in the standard node/link method of network definition with intersections represented by nodes only. Such problem could be solved by expansion technique adding extra links and nodes to the network for describing turn penalties, but this method could not apply to large networks as well as dynamic case due to its overwhelming additional works. This paper proposes a link-based shortest path algorithm for the travel information in real road network where exists turn prohibitions. It enables to provide efficient alternative paths under consideration of overlaps among paths. The algorithm builds each path based on the degree of overlapping between each path and stops building new path when the degree of overlapping ratio exceeds its criterion. Because proposed algorithm builds the shortest path based on the link-end cost instead or node cost and constructs path between origin and destination by link connection, the network expansion does not require. Thus it is possible to save the time or network modification and of computer running. Some numerical examples are used for test of the model proposed in the paper.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.39 no.9
• /
• pp.1-11
• /
• 2002

The optimal AP placement for wireless LAN is important factor for improving service quality and reducing cost. Decision of AP placement is depend on radio signal strength, environment factor and logical area property, which is user's frequently posed place. This paper proposes optimal multiple AP placement method based on radio prediction tool. The proposed method can get flexibility in multiple AP placement using user defined parameter and the optimization design uses Hopfield network algorithm. And path-loss model is used for one of radion prediction model. The result of simulation shows that it is efficiently reduces the process to find optimal AP placement. And the proposed optimization design of multiple AP placement can improve service quality for wireless LAN.

• Journal of the Korean Geographical Society
• /
• v.43 no.2
• /
• pp.220-234
• /
• 2008

In many security monitoring contexts, the performance or efficiency of surveillance sensors/networks based on a single sensor type may be limited by environmental conditions, like illumination change. It is well known that different modes of sensors can be complementary, compensating for failures or limitations of individual sensor types. From a location analysis and modeling perspective, a challenge is how to locate different modes of sensors to support security monitoring. A coverage-based optimization model is proposed as a way to simultaneously site k different sensor types. This model considers common coverage among different sensor types as well as overlapping coverage for individual sensor types. The developed model is used to site sensors in an urban area. Computational results show that common and overlapping coverage can be modeled simultaneously, and a rich set of solutions exists reflecting the tradeoff between common and overlapping coverage.

• Korean Journal of Remote Sensing
• /
• v.33 no.4
• /
• pp.423-436
• /
• 2017

This paper presents a deep learning-based road segmentation framework from very high-resolution orthophotos. The proposed method uses Deep Convolutional Autoencoders for end-to-end mapping of orthophotos to road segmentations. In addition, a set of post-processing steps were applied to make the model outputs GIS-ready data that could be useful for various applications. The optimization of the model's parameters is explained which was conducted via grid search method. The model was trained and implemented in Keras, a high-level deep learning framework run on top of Tensorflow. The results show that the proposed model with the best-obtained hyperparameters could segment road objects from orthophotos at an average accuracy of 88.5%. The results of optimization revealed that the best optimization algorithm and activation function for the studied task are Stochastic Gradient Descent (SGD) and Exponential Linear Unit (ELU), respectively. In addition, the best numbers of convolutional filters were found to be 8 for the first and second layers and 128 for the third and fourth layers of the proposed network architecture. Moreover, the analysis on the time complexity of the model showed that the model could be trained in 4 hours and 50 minutes on 1024 high-resolution images of size $106{\times}106pixels$, and segment road objects from similar size and resolution images in around 14 minutes. The results show that the deep learning models such as Convolutional Autoencoders could be a best alternative to traditional machine learning models for road segmentation from aerial photographs.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.7
• /
• pp.2389-2413
• /
• 2015

High data rates in long-term evolution (LTE) networks can affect the mobility of networks and their performance. The speed and motion of user equipment (UE) can compromise seamless connectivity. However, a proper handover (HO) decision can maintain quality of service (QoS) and increase system throughput. While this may lead to an increase in complexity and operational costs, self-optimization can enhance network performance by improving resource utilization and user experience and by reducing operational and capital expenditure. In this study, we propose the self-optimization of HO parameters based on fuzzy logic control (FLC) and multiple preparation (MP), which we name FuzAMP. Fuzzy logic control can be used to control self-optimized HO parameters, such as the HO margin and time-to-trigger (TTT) based on multiple criteria, viz HO ping pong (HOPP), HO failure (HOF) and UE speeds. A MP approach is adopted to overcome the hard HO (HHO) drawbacks, such as the large delay and unreliable procedures caused by the break-before-make process. The results of this study show that the proposed method significantly reduces HOF, HOPP, and packet loss ratio (PLR) at various UE speeds compared to the HHO and the enhanced weighted performance HO parameter optimization (EWPHPO) algorithms.

• The Journal of the Korea Contents Association
• /
• v.7 no.8
• /
• pp.59-65
• /
• 2007

Overlay Multicast is an effective method for efficient utilization of system resources and network bandwidth without a need for hardware customization. Multicast tree reconstruction is required when a non-leaf node leaves or fails. However frequent multicast tree reconstruction introduces serious degradation in performance. In this paper, we propose a tree performance optimization algorithm to solve this defect by using information(RTCP-probing) that becomes a periodic feedback to a source node from each child node. The proposed model is a mechanism performed when a parent node seems to cause deterioration in the tree performance. We have improved the performance of the whole service tree using the mechanism and hence composing an optimization tree. The simulation results show that our proposal stands to be an effective method that can be applied to not only the proposed model but also to existing techniques.