• Journal of the Korea Convergence Society
• /
• v.8 no.7
• /
• pp.29-37
• /
• 2017

Compared to the commercial networks, R&E networks have the strength such as flexible network engineering and design. Based on those features of R&E networks, we propose our storage cloud service model which supports general-purpose network users in a central region and experimental network users in distributed regions simultaneously. We prototype our service model utilizing multiple proxy controllers of OpenStack Swift service in order to deploy several regions via experimental backbone networks. Our experiments on the influence of the network latency and the size of data to be transmitted show that the bigger size of data is preferable to the smaller size of data in an experimental backbone network where the network latency increases within 10ms because the rate of throughput decline in the bigger object is comparatively small. It means that our service model is appropriate for experimental network users who directly access the service in order to move intermittently high volume of data as well as normal users in the central region who access the service frequently.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.3
• /
• pp.381-388
• /
• 2022

In instance segmentation, Mask-RCNN is mostly used as a base model. Increasing the performance of Mask-RCNN is meaningful because it affects the performance of the derived model. Mask-RCNN has a transform module for unifying size of input images. In this paper, to improve the Mask-RCNN, we apply deep-learning-based ASSR to the resizing part in the transform module and inject calculated scale information into the model using IM(Integration Module). The proposed IM improves instance segmentation performance by 2.5 AP higher than Mask-RCNN in the COCO dataset, and in the periment for optimizing the IM location, the best performance was shown when it was located in the 'Top' before FPN and backbone were combined. Therefore, the proposed method can improve the performance of models using Mask-RCNN as a base model.

• Journal of the Korean Geotechnical Society
• /
• v.40 no.3
• /
• pp.33-39
• /
• 2024

Inspecting cracks to determine a structure's condition is crucial for accurate safety diagnosis. However, visual crack inspection methods can be subjective and are dependent on field conditions, thereby resulting in low reliability. To address this issue, this study automates the detection of concrete cracks in image data using ResNet, FPN, and the Mask R-CNN components as the backbone, neck, and head of a convolutional neural network. The performance of the proposed model is analyzed using the intersection over the union (IoU). The experimental dataset contained 1,203 images divided into training (70%), validation (20%), and testing (10%) sets. The model achieved an IoU value of 95.83% for testing, and there were no cases where the crack was not detected. These findings demonstrate that the proposed model realized highly accurate detection of concrete cracks in image data.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.19 no.5
• /
• pp.178-188
• /
• 2020

This paper proposes a lane recognition CNN network using split-attention network as a backbone to extract feature. Split-attention is a method of assigning weight to each channel of a feature map in the CNN feature extraction process; it can reliably extract the features of an image during the rapidly changing driving environment of a vehicle. The proposed deep neural networks in this paper were trained and evaluated using the Tusimple data set. The change in performance according to the number of layers of the backbone network was compared and analyzed. A result comparable to the latest research was obtained with an accuracy of up to 96.26, and FN showed the best result. Therefore, even in the driving environment of an actual vehicle, stable lane recognition is possible without misrecognition using the model proposed in this study.

• Journal of Advanced Research in Ocean Engineering
• /
• v.4 no.4
• /
• pp.174-184
• /
• 2018

We herein propose a new design procedure of a flexible container ship model where the vertical bending and torsional vibration modes are similar to its prototype. To achieve similarity in torsional vibration mode shapes, the height of the shear center of the model must be located below the bottom hull, similar to an actual container ship with large opening decks. Therefore, we designed a ship model by imparting appropriate stiffness to the hull, using urethane foam without a backbone. We built a container ship model according to this design strategy and validated its dynamic elastic properties using a decay test. We measured wave-induced structural vibrations and present the results of tank experiments in regular and freak waves.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.05a
• /
• pp.2-4
• /
• 2023

Sequence data plays an important role in the field of intelligence, especially for industrial control, traffic control and other aspects. Finding abnormal parts in sequence data has long been an application field of AI technology. In this paper, we propose an anomaly detection method for sequence data using a diffusion model. The diffusion model has two major advantages: interpretability derived from rigorous mathematical derivation and unrestricted selection of backbone models. This method uses the diffusion model to predict and reconstruct the sequence data, and then detects the abnormal part by comparing with the real data. This paper successfully verifies the feasibility of the diffusion model in the field of anomaly detection. We use the combination of MLP and diffusion model to generate data and compare the generated data with real data to detect anomalous points.

• Land and Housing Review
• /
• v.14 no.2
• /
• pp.125-135
• /
• 2023

Drones are used in various fields, including land survey, transportation, forestry/agriculture, marine, environment, disaster prevention, water resources, cultural assets, and construction, as their industrial importance and market size have increased. In this study, image data for deep learning was collected using a mavic3 drone capturing images at a shooting altitude was 20 m with ×7 magnification. Swin Transformer and UperNet were employed as the backbone and architecture of the deep learning model. About 800 sheets of labeled data were augmented to increase the amount of data. The learning process encompassed three rounds. The Cross-Entropy loss function was used in the first and second learning; the Tversky loss function was used in the third learning. In the future, when the crack detection model is advanced through convergence with the Internet of Things (IoT) through additional research, it will be possible to detect patching or potholes. In addition, it is expected that real-time detection tasks of drones can quickly secure the detection of pavement maintenance sections.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.2
• /
• pp.680-699
• /
• 2015

The use of web service has been increased rapidly, with an increase in the number of available services, finding the exact service is the challenging task. Service discovery is the most significant job to complete the service discoverers needs. In order to achieve the efficient service discovery, we focus on designing a cluster based service discovery model for service registering and service provisioning among all mobile nodes in a mobile ad hoc network (MANETs). A dynamic backbone of nodes (i.e. cluster heads) that forms a service repository to which MANET nodes can publish their services and/or send their service queries. The designed model is based on storing services with their service description on cluster head nodes that are found in accordance with the proposed cluster head election model. In addition to identifying and analyzing the system parameters for finding the effectiveness of our model, this paper studies the stability analysis of the network, overhead of the cluster, and bandwidth utilization and network traffic is evaluated using analytic derivations and experimental evaluation has been done.

• Computers and Concrete
• /
• v.20 no.1
• /
• pp.85-97
• /
• 2017

With a special attention to the different stages of a typical loading path travelled in a fluid confined concrete test, this paper introduces a physically consistent model for the stress-strain curve of actively confined normal-strength concrete in the axial direction. The model comprises of the five elements of: (1) a criterion for the peak or failure strength, (2) an equation for the peak strain, (3) a backbone hydrostatic curve, (4) a transient hardening curve linking the point of departure from the hydrostatic curve to the failure point, and finally (5) a set of formulas for the post-peak region. Alongside, relevant details and shortcomings of existing models will be discussed in each part. Finally, the accuracy and efficiency of the proposed model have been verified in a set of simulations which compare well with the experimental results from the literature.

• Smart Media Journal
• /
• v.9 no.3
• /
• pp.59-70
• /
• 2020

This paper presents a fully automatic tool to recognize the liver region from CT images based on a deep learning model, namely Multiple Filter U-net, MFUnet. The advantages of both U-net and Multiple Filters were utilized to construct an autoencoder model, called MFUnet for segmenting the liver region from computed tomograph. The MFUnet architecture includes the autoencoding model which is used for regenerating the liver region, the backbone model for extracting features which is trained on ImageNet, and the predicting model used for liver segmentation. The LiTS dataset and Chaos dataset were used for the evaluation of our research. This result shows that the integration of Multiple Filter to U-net improves the performance of liver segmentation and it opens up many research directions in medical imaging processing field.