• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.13 no.5
• /
• pp.313-322
• /
• 2020

A large volume of data will improve the robustness of deep learning models and avoid overfitting problems. In automatic tongue segmentation, the availability of annotated tongue images is often limited because of the difficulty of collecting and labeling the tongue image datasets in reality. Data augmentation can expand the training dataset and increase the diversity of training data by using label-preserving transformations without collecting new data. In this paper, augmented tongue image datasets were developed using seven augmentation techniques such as image cropping, rotation, flipping, color transformations. Performance of the data augmentation techniques were studied using state-of-the-art transfer learning models, for instance, InceptionV3, EfficientNet, ResNet, DenseNet and etc. Our results show that geometric transformations can lead to more performance gains than color transformations and the segmentation accuracy can be increased by 5% to 20% compared with no augmentation. Furthermore, a random linear combination of geometric and color transformations augmentation dataset gives the superior segmentation performance than all other datasets and results in a better accuracy of 94.98% with InceptionV3 models.

• Structural Engineering and Mechanics
• /
• v.85 no.4
• /
• pp.469-484
• /
• 2023

A deep recursive bidirectional Cuda Deep Neural Network Long Short Term Memory (Bi-CuDNNLSTM) layer is recruited in this paper to predict the entire force time histories, and the corresponding hysteresis and backbone curves of reinforced concrete (RC) bridge piers using experimental fast and slow cyclic tests. The proposed stacked Bi-CuDNNLSTM layers involve multiple uncertain input variables, including horizontal actuator displacements, vertical actuators axial loads, the effective height of the bridge pier, the moment of inertia, and mass. The functional application programming interface in the Keras Python library is utilized to develop a deep learning model considering all the above various input attributes. To have a robust and reliable prediction, the dataset for both the fast and slow cyclic tests is split into three mutually exclusive subsets of training, validation, and testing (unseen). The whole datasets include 17 RC bridge piers tested experimentally ten for fast and seven for slow cyclic tests. The results bring to light that the mean absolute error, as a loss function, is monotonically decreased to zero for both the training and validation datasets after 5000 epochs, and a high level of correlation is observed between the predicted and the experimentally measured values of the force time histories for all the datasets, more than 90%. It can be concluded that the maximum mean of the normalized error, obtained through Box-Whisker plot and Gaussian distribution of normalized error, associated with unseen data is about 10% and 3% for the fast and slow cyclic tests, respectively. In recapitulation, it brings to an end that the stacked Bi-CuDNNLSTM layer implemented in this study has a myriad of benefits in reducing the time and experimental costs for conducting new fast and slow cyclic tests in the future and results in a fast and accurate insight into hysteretic behavior of bridge piers.

• International Journal of Contents
• /
• v.16 no.2
• /
• pp.8-18
• /
• 2020

Data is critical in deep learning but the scarcity of data often occurs in research, especially in the preparation of the paired training data. In this paper, document image binarization with unpaired data is studied by introducing adversarial learning, excluding the need for supervised or labeled datasets. However, the simple extension of the previous unpaired training to binarization inevitably leads to poor performance compared to paired data training. Thus, a new deep learning approach is proposed by introducing a multi-diversity of higher quality generated images. In this paper, a two-stage model is proposed that comprises the generative adversarial network (GAN) followed by the U-net network. In the first stage, the GAN uses the unpaired image data to create paired image data. With the second stage, the generated paired image data are passed through the U-net network for binarization. Thus, the trained U-net becomes the binarization model during the testing. The proposed model has been evaluated over the publicly available DIBCO dataset and it outperforms other techniques on unpaired training data. The paper shows the potential of using unpaired data for binarization, for the first time in the literature, which can be further improved to replace paired data training for binarization in the future.

• Asia-Pacific Journal of Business
• /
• v.12 no.4
• /
• pp.241-251
• /
• 2021

Purpose - Drawing on the content-process approach, this study examines the effect of employees' perceived utility of education and training in small and medium enterprises (SMEs) on their satisfaction. In addition, this study investigates how the human resource management department' activities moderate the relationship between employees' perceived utility of education and training and satisfaction. Design/methodology/approach - This study predicts the positive relationship between employees' perceived utility of education and training and satisfaction, and HR activities strengthens this positive relationship. To test these hypotheses, this study utilized Human Capital Corporate Panel (HCCP) datasets, especially 2017 data at the individual level. The number of the final sample is 425 for the test. Moreover, this study used the hierarchical regression model with SPSS. Finding - As predicted, the analytical results with the hierarchical regression model showed that employees' percieved utility of education and training and satisfaction were positively related. In addition, HR activities strengthened this relationship between employees' percieved utility of education and training and satisfaction. Research implications or Originality - This study will provide academic and practical implications for future research on human resource development, especially SMEs by deepening an understanding of the important factors in order to increase employees' satisfaction of education and training. the number of viewers is found in most American films released in Korea.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.11
• /
• pp.29-42
• /
• 2023

In this paper, we propose a method to enhance the prediction accuracy of solar irradiance for three major South Korean cities: Seoul, Busan, and Incheon. Our method entails the development of five generative models-vanilla GAN, CTGAN, Copula GAN, WGANGP, and TVAE-to generate independent variables that mimic the patterns of existing training data. To mitigate the bias in model training, we derive values for the dependent variables using random forests and deep neural networks, enriching the training datasets. These datasets are integrated with existing data to form comprehensive solar irradiance prediction models. The experimentation revealed that the augmented datasets led to significantly improved model performance compared to those trained solely on the original data. Specifically, CTGAN showed outstanding results due to its sophisticated mechanism for handling the intricacies of multivariate data relationships, ensuring that the generated data are diverse and closely aligned with the real-world variability of solar irradiance. The proposed method is expected to address the issue of data scarcity by augmenting the training data with high-quality synthetic data, thereby contributing to the operation of solar power systems for sustainable development.

• IEIE Transactions on Smart Processing and Computing
• /
• v.6 no.3
• /
• pp.183-192
• /
• 2017

k-nearest neighbor (K-NN) is a well-known classification algorithm, being feature space-based on nearest-neighbor training examples in machine learning. However, K-NN, as we know, is a lazy learning method. Therefore, if a K-NN-based system very much depends on a huge amount of history data to achieve an accurate prediction result for a particular task, it gradually faces a processing-time performance-degradation problem. We have noticed that many researchers usually contemplate only classification accuracy. But estimation speed also plays an essential role in real-time prediction systems. To compensate for this weakness, this paper proposes correlation coefficient-based clustering (CCC) aimed at upgrading the performance of K-NN by leveraging processing-time speed and plurality rule-based density (PRD) to improve estimation accuracy. For experiments, we used real datasets (on breast cancer, breast tissue, heart, and the iris) from the University of California, Irvine (UCI) machine learning repository. Moreover, real traffic data collected from Ojana Junction, Route 58, Okinawa, Japan, was also utilized to lay bare the efficiency of this method. By using these datasets, we proved better processing-time performance with the new approach by comparing it with classical K-NN. Besides, via experiments on real-world datasets, we compared the prediction accuracy of our approach with density peaks clustering based on K-NN and principal component analysis (DPC-KNN-PCA).

• International Journal of Computer Science & Network Security
• /
• v.22 no.9
• /
• pp.23-30
• /
• 2022

Glaucoma is a chronic neuropathy that affects the optic nerve which can lead to blindness. The detection and prediction of glaucoma become possible using deep neural networks. However, the detection performance relies on the availability of a large number of data. Therefore, we propose different frameworks, including a hybrid of a generative adversarial network and a convolutional neural network to automate and increase the performance of glaucoma detection. The proposed frameworks are evaluated using five public glaucoma datasets. The framework which uses a Deconvolutional Generative Adversarial Network (DCGAN) and a DenseNet pre-trained model achieves 99.6%, 99.08%, 99.4%, 98.69%, and 92.95% of classification accuracy on RIMONE, Drishti-GS, ACRIMA, ORIGA-light, and HRF datasets respectively. Based on the experimental results and evaluation, the proposed framework closely competes with the state-of-the-art methods using the five public glaucoma datasets without requiring any manually preprocessing step.

• Korean Journal of Materials Research
• /
• v.33 no.7
• /
• pp.273-278
• /
• 2023

In this study, based on the saturation magnetic flux density experimental values (B_s) of 622 Fe-based bulk metallic glasses (BMGs), regression models were applied to predict B_s using artificial neural networks (ANN), and prediction performance was evaluated. Model performance evaluation was investigated by using the F1 score together with the coefficient of determination (R² score), which is mainly used in regression models. The coefficient of determination can be used as a performance indicator, since it shows the predicted results of the saturation magnetic flux density of full material datasets in a balanced way. However, the BMG alloy contains iron and requires a high saturation magnetic flux density to have excellent applicability as a soft magnetic material, and in this study F1 score was used as a performance indicator to better predict B_s above the threshold value of B_s (1.4 T). After obtaining two ANN models optimized for the R² and F1 score conditions, respectively, their prediction performance was compared for the test data. As a case study to evaluate the prediction performance, new Fe-based BMG datasets that were not included in the training and test datasets were predicted using the two ANN models. The results showed that the model with an excellent F1 score achieved a more accurate prediction for a material with a high saturation magnetic flux density.

• The Journal of Information Systems
• /
• v.32 no.4
• /
• pp.119-136
• /
• 2023

Purpose We introduce a groundbreaking surveillance system explicitly designed to overcome the limitations typically associated with conventional surveillance systems, which often focus primarily on object-centric behavior analysis. Design/methodology/approach The study introduces an innovative approach to risk assessment in surveillance, employing image captioning to generate descriptive captions that effectively encapsulate the interactions among objects, actions, and spatial elements within observed scenes. To support our methodology, we developed a distinctive dataset comprising pairs of [image-caption-danger score] for training purposes. We fine-tuned the BLIP-2 model using this dataset and utilized BERT to decipher the semantic content of the generated captions for assessing risk levels. Findings In a series of experiments conducted with our self-constructed datasets, we illustrate that these datasets offer a wealth of information for risk assessment and display outstanding performance in this area. In comparison to models pre-trained on established datasets, our generated captions thoroughly encompass the necessary object attributes, behaviors, and spatial context crucial for the surveillance system. Additionally, they showcase adaptability to novel sentence structures, ensuring their versatility across a range of contexts.

• Journal of Broadcast Engineering
• /
• v.26 no.7
• /
• pp.855-861
• /
• 2021

Neural Architecture Search (NAS) is cutting-edge technology in the machine learning community. NAS Without Training (NASWOT) recently has been proposed to tackle the high demand of computational resources in NAS by leveraging some indicators to predict the performance of architectures before training. The advantage of these indicators is that they do not require any training. Thus, NASWOT reduces the searching time and computational cost significantly. However, NASWOT only considers high-performing networks which does not guarantee a fast inference speed on hardware devices. In this paper, we propose a multi objectives reward function, which considers the network's latency and the predicted performance, and incorporate it into the Reinforcement Learning approach to search for the best networks with low latency. Unlike other methods, which use FLOPs to measure the latency that does not reflect the actual latency, we obtain the network's latency from the hardware NAS bench. We conduct extensive experiments on NAS-Bench-201 using CIFAR-10, CIFAR-100, and ImageNet-16-120 datasets, and show that the proposed method is capable of generating the best network under latency constrained without training subnetworks.