• Smart Structures and Systems
• /
• v.33 no.6
• /
• pp.449-463
• /
• 2024

Automated crack detection is crucial for structural health monitoring and post-earthquake rapid damage detection. However, realizing high precision automatic crack detection in the absence of corresponding manual labeling presents a formidable challenge. This paper presents a novel crack segmentation transfer learning method and a novel crack segmentation model called Swin-CrackFormer. The proposed method facilitates efficient crack image style transfer through a meticulously designed data preprocessing technique, followed by the utilization of a GAN model for image style transfer. Moreover, the proposed Swin-CrackFormer combines the advantages of Transformer and convolution operations to achieve effective local and global feature extraction. To verify the effectiveness of the proposed method, this study validates the proposed method on three unlabeled crack datasets and evaluates the Swin-CrackFormer model on the METU dataset. Experimental results demonstrate that the crack transfer learning method significantly improves the crack segmentation performance on unlabeled crack datasets. Moreover, the Swin-CrackFormer model achieved the best detection result on the METU dataset, surpassing existing crack segmentation models.

• Information Systems Review
• /
• v.12 no.3
• /
• pp.19-48
• /
• 2010

This study focused on characterizing various factors in order for learners to maintain their interests in learning and to maximize learning effects as the top priority purpose of university e-Learning, on the basis of results of conceptual studies on existing e-Learning and practical studies, and then on examining them practically. It also analyzed which factors would have greater influence on learning effects of e-Learning in general. Moreover, in comparison with existing numerous studies which examined only factor such as learning effects of e-Learning, it analyzed such things in detail according to division into three items such as learning satisfaction, learning transfer and learning recommendation. To achieve such purposes of the study, it characterized and set 3 factors such as learning contents, instructional design and user convenience on the assumption that such factors have a significant influence on learning effects of e-Learning. Moreover, the factor of learning contents includes 3 detailed elements, i.e., learning issue and objective, knowledge information, and consistency and propriety, and the factor of instructional design includes 4 detailed elements, i.e., interest and sympathy, interaction, contents presentation and explanatory strategy. Lastly, the factor of user convenience includes 2 detailed elements such as screen configuration, and check-up of contents and teaching schedule. According to analytical results, it showed all 3 factors such as learning contents, instructional design and user convenience have a significant influence on learning effects of e-Learning(i.e., learning satisfaction, learning transfer and learning recommendation). In more detail, it showed the learning issue and objective from the factor of learning contents have the greatest influence on learning satisfaction of e-Learning. Then, it is the most important to set the learning issue and objective with given priority to learners and set the learning objective estimable, in order to raise the learning satisfaction. It showed the contents presentation from the factor of instructional design on the learning transfer. Therefore, it is the most important to structuralize mutual relation and presentation orders to promote learning systematically and to let learners access to such things, for the purpose of raising the learning transfer. Moreover, it showed the interest and sympathy from the factor of instructional design has the greatest influence on the learning recommendation. Thus, it is the most important to promote learners' interests to the maximum using well-timed media, and to give a lecture enough to arouse learners' sympathy.

• Journal of Internet Computing and Services
• /
• v.20 no.4
• /
• pp.91-102
• /
• 2019

In autonomous driving systems, the ability to classify pedestrians in images captured by cameras is very important for pedestrian safety. In the past, after extracting features of pedestrians with HOG(Histogram of Oriented Gradients) or SIFT(Scale-Invariant Feature Transform), people classified them using SVM(Support Vector Machine). However, extracting pedestrian characteristics in such a handcrafted manner has many limitations. Therefore, this paper proposes a method to classify pedestrians reliably and effectively using CNN's(Convolutional Neural Network) deep features and transfer learning. We have experimented with both the fixed feature extractor and the fine-tuning methods, which are two representative transfer learning techniques. Particularly, in the fine-tuning method, we have added a new scheme, called M-Fine(Modified Fine-tuning), which divideslayers into transferred parts and non-transferred parts in three different sizes, and adjusts weights only for layers belonging to non-transferred parts. Experiments on INRIA Person data set with five CNN models(VGGNet, DenseNet, Inception V3, Xception, and MobileNet) showed that CNN's deep features perform better than handcrafted features such as HOG and SIFT, and that the accuracy of Xception (threshold = 0.5) isthe highest at 99.61%. MobileNet, which achieved similar performance to Xception and learned 80% fewer parameters, was the best in terms of efficiency. Among the three transfer learning schemes tested above, the performance of the fine-tuning method was the best. The performance of the M-Fine method was comparable to or slightly lower than that of the fine-tuningmethod, but higher than that of the fixed feature extractor method.

• The Journal of the Acoustical Society of Korea
• /
• v.35 no.2
• /
• pp.143-148
• /
• 2016

Deep neural network that effectively capture the characteristics of data has been widely used in various applications. However, the amount of sound database is often insufficient for learning the deep neural network properly, so resulting in overfitting problems. In this paper, we propose a transfer learning framework that can effectively train the deep neural network even with insufficient sound event data by employing rich speech or music data. A series of experimental results verify that proposed method performs significantly better than the baseline deep neural network that was trained only with small sound event data.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.1
• /
• pp.613-618
• /
• 2023

In this paper, a monitoring program that can automatically determine whether a patient admitted to an isolation room acts out of a stable state through a screen photographed in real time is described. The motion recognition model of this program was built by learning through transfer learning. 900 images were used for the three movements, and this program was developed for the web to support all environments. The model was determined with high accuracy to determine the state of the subject hospitalized in the isolation room, and can be applied by applying it to the existing isolation room monitoring system.

• Nuclear Engineering and Technology
• /
• v.54 no.12
• /
• pp.4751-4758
• /
• 2022

Two-phase flow may almost exist in every branch of the energy industry. For the corresponding engineering design, it is very essential and crucial to monitor flow patterns and their transitions accurately. With the high-speed development and success of deep learning based on convolutional neural network (CNN), the study of flow pattern identification recently almost focused on this methodology. Additionally, the photographing technique has attractive implementation features as well, since it is normally considerably less expensive than other techniques. The development of such a two-phase flow pattern online monitoring system is the objective of this work, which seldom studied before. The ongoing preliminary engineering design (including hardware and software) of the system are introduced. The flow pattern identification method based on CNNs and transfer learning was discussed in detail. Several potential CNN candidates such as ALexNet, VggNet16 and ResNets were introduced and compared with each other based on a flow pattern dataset. According to the results, ResNet50 is the most promising CNN network for the system owing to its high precision, fast classification and strong robustness. This work can be a reference for the online monitoring system design in the energy system.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.5
• /
• pp.117-122
• /
• 2024

Many studies on software defect prediction have been conducted, but it has been difficult to use them due to a lack of training data. Cross-project defect prediction is a technique to solve this problem, where a prediction model learned with sufficient training data from existing source project is used to predict defects in the target project. Before learning, domain adaptation techniques, a type of transfer learning, are used to minimize the difference in data distribution between the two projects. In this paper, we produced new prediction models using W-BDA and MEDA and compared their performance with existing models using TCA and BDA. As a result of the evaluation experiment, MEDA showed irregular and poor performance compared to other models, but BDA showed better performance than TCA, and W-BDA showed slightly better performance than BDA.

• Journal of the Institute of Convergence Signal Processing
• /
• v.24 no.1
• /
• pp.34-41
• /
• 2023

In recent years, as the range of human activities has increased, the introduction of alien species has become frequent. Among them, raccoons have been designated as harmful animals since 2020. Raccoons are similar in size and shape to raccoon dogs, so they generally need to be distinguished in capturing them. To solve this problem, we use VGG19, ResNet152V2, InceptionV3, InceptionResNet and NASNet, which are CNN deep learning models specialized for image classification. The parameters to be used for learning are pre-trained with a large amount of data, ImageNet. In order to classify the raccoon and raccoon dog datasets as outward features of animals, the image was converted to grayscale and brightness was normalized. Augmentation methods were applied using left and right inversion, rotation, scaling, and shift to create sufficient data for transfer learning. The FCL consists of 1 layer for the non-augmented dataset while 4 layers for the augmented dataset. Comparing the accuracy of various augmented datasets, the performance increased as more augmentation methods were applied.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.17 no.4
• /
• pp.213-220
• /
• 2024

This paper proposes an optimization strategy for performing Federated Learning between devices and edge clouds using WebAssembly (WASM). The proposed strategy aims to maximize efficiency by conducting partial training on devices and the remaining training on edge clouds. Specifically, it mathematically describes and evaluates methods to optimize data transfer between GPU memory segments and the overlapping of computational tasks to reduce overall training time and improve GPU utilization. Through various experimental scenarios, we confirmed that asynchronous data transfer and task overlap significantly reduce training time, enhance GPU utilization, and improve model accuracy. In scenarios where all optimization techniques were applied, training time was reduced by 47%, GPU utilization improved to 91.2%, and model accuracy increased to 89.5%. These results demonstrate that asynchronous data transfer and task overlap effectively reduce GPU idle time and alleviate bottlenecks. This study is expected to contribute to the performance optimization of Federated Learning systems in the future.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.19 no.1
• /
• pp.32-43
• /
• 2020

The transfer time is an important factor in establishing public transportation planning and policy. Therefore, in this study, the influencing factors of the transfer time for transit users were identified using smart card data, and the estimation results for the transfer time using the deep learning method such as deep neural network models were compared with traditional regression models. First, the intervals and the distance to the bus stop had positive effects on the subway-to-bus transfer time, and the number of bus routes had a negative effect. This also showed that the transfer time is affected by the area in which the subway station exists. Based on the influencing factors of the transfer time, the deep learning models were developed and their estimation results were compared with the regression model. For model performance, the deep learning models were better than those of the regression models. These results can be used as basic data for transfer policies such as the differential application of transit allowance times according to region.