• International Journal of Internet, Broadcasting and Communication
• /
• v.11 no.4
• /
• pp.31-36
• /
• 2019

Unsupervised neural networks have not caught enough attention until Generative Adversarial Network (GAN) was proposed. By using both the generator and discriminator networks, GAN can extract the main characteristic of the original dataset and produce new data with similarlatent statistics. However, researchers understand fully that training GAN is not easy because of its unstable condition. The discriminator usually performs too good when helping the generator to learn statistics of the training datasets. Thus, the generated data is not compelling. Various research have focused on how to improve the stability and classification accuracy of GAN. However, few studies delve into how to improve the training efficiency and to save training time. In this paper, we propose a novel optimizer, named FAST-ADAM, which integrates the Lookahead to ADAM optimizer to train the generator of a semi-supervised generative adversarial network (SSGAN). We experiment to assess the feasibility and performance of our optimizer using Canadian Institute For Advanced Research - 10 (CIFAR-10) benchmark dataset. From the experiment results, we show that FAST-ADAM can help the generator to reach convergence faster than the original ADAM while maintaining comparable training accuracy results.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2019.05a
• /
• pp.338-340
• /
• 2019

Recently, the speed and size of data accumulation are increasing due to the development of networks. There are many difficulties in classifying these data. One of the difficulties is the difficulty of labeling. Labeling is usually done by people, but it is very difficult for everyone to understand the data in the same way and it is very difficult to label them on the same basis. In order to solve this problem, we implemented GAN to generate new image based on input image and to learn input data indirectly by using it for learning. This suggests that the accuracy of classification can be increased by increasing the number of learning data.

• Journal of Ocean Engineering and Technology
• /
• v.34 no.6
• /
• pp.442-453
• /
• 2020

Underwater color images suffer from low visibility and color cast effects caused by light attenuation by water and floating particles. This study applied single image enhancement techniques to enhance the quality of underwater images and compared their performance with real underwater images taken in Korean waters. Dark channel prior (DCP), gradient transform, image fusion, and generative adversarial networks (GAN), such as cycleGAN and underwater GAN (UGAN), were considered for single image enhancement. Their performance was evaluated in terms of underwater image quality measure, underwater color image quality evaluation, gray-world assumption, and blur metric. The DCP saturated the underwater images to a specific greenish or bluish color tone and reduced the brightness of the background signal. The gradient transform method with two transmission maps were sensitive to the light source and highlighted the region exposed to light. Although image fusion enabled reasonable color correction, the object details were lost due to the last fusion step. CycleGAN corrected overall color tone relatively well but generated artifacts in the background. UGAN showed good visual quality and obtained the highest scores against all figures of merit (FOMs) by compensating for the colors and visibility compared to the other single enhancement methods.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.7
• /
• pp.27-33
• /
• 2022

In this paper, we propose a novel generator architecture for multi-domain style transfer method not an image to image translation, as a method of generating a styled image by transfering a style to the content image. A latent vector and Gaussian noises are added to the generator of GAN so that a high quality image is generated while considering the characteristics of various data distributions for each domain and preserving the features of the content data. With the generator architecture of the proposed GAN, networks are configured and presented so that the content image can learn the styles for each domain well, and it is applied to the domain composed of images of the four seasons to show the high resolution style transfer results.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.25 no.5
• /
• pp.476-486
• /
• 2022

Synthetic dynamic infrared image generation from the given virtual environment is being the primary goal to simulate the output of the infra-red(IR) camera installed on a vehicle to evaluate the control algorithm for various search & reconnaissance missions. Due to the difficulty to obtain actual IR data in complex environments, Artificial intelligence(AI) has been used recently in the field of image data generation. In this paper, CycleGAN technique is applied to obtain a more realistic synthetic IR image. We added the Structural Similarity Index Measure(SSIM) loss function to the L1 loss function to generate a more realistic synthetic IR image when the CycleGAN image is generated. From the simulation, it is applicable to the guided-missile flight simulation tests by using the synthetic infrared image generated by the proposed technique.

• 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.

• Communications for Statistical Applications and Methods
• /
• v.30 no.6
• /
• pp.589-603
• /
• 2023

Deep generative models target to infer the underlying true data distribution, and it leads to a huge success in generating fake-but-realistic data. Regarding such a perspective, the data attributes can be a crucial factor in the data generation process since non-existent counterfactual samples can be generated by altering certain factors. For example, we can generate new portrait images by flipping the gender attribute or altering the hair color attributes. This paper proposes counterfactual disentangled variational autoencoder generative adversarial networks (CDVAE-GAN), specialized for data attribute level counterfactual data generation. The structure of the proposed CDVAE-GAN consists of variational autoencoders and generative adversarial networks. Specifically, we adopt a Gaussian variational autoencoder to extract low-dimensional disentangled data features and auxiliary Bernoulli latent variables to model the data attributes separately. Also, we utilize a generative adversarial network to generate data with high fidelity. By enjoying the benefits of the variational autoencoder with the additional Bernoulli latent variables and the generative adversarial network, the proposed CDVAE-GAN can control the data attributes, and it enables producing counterfactual data. Our experimental result on the CelebA dataset qualitatively shows that the generated samples from CDVAE-GAN are realistic. Also, the quantitative results support that the proposed model can produce data that can deceive other machine learning classifiers with the altered data attributes.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.6
• /
• pp.622-628
• /
• 2020

This paper proposes a music generation technique that automatically generates trot music using a Generative Adversarial Network (GAN) model composed of a Recurrent Neural Network (RNN). The proposed method uses a method of creating a chord as a skeleton of the music, creating a melody and bass in stages based on the chord progression made, and attaching it to the corresponding chord to complete the structured piece. Also, a new chorus chord progression is created from the verse chord progression by applying the characteristics of a trot song that repeats the structure divided into an individual section, such as intro, verse, and chorus. And it extends the length of the created trot. The quality of the generated music was specified using subjective evaluation and objective evaluation methods. It was confirmed that the generated music has similar characteristics to the existing trot.

• The Journal of Korean Medical History
• /
• v.23 no.2
• /
• pp.81-87
• /
• 2010

1. An analysis of "GuGeupGanIBang(救急簡易方)" through Gwakran(霍亂) showed that the reference books are not concretely cited, but the symptoms are described in detail ed to let the common people understand and deal with Gwakran(霍亂) easily. It has the most detailed among Korean annotation emergency treatment books. It focuses on moxibustion and hot medicinal compress rather than acupuncture. The Korean annotation is translated in the order of the Korean language, so anyone can execute the treatments. The annotation used a lot of active verbs, so it is more of a book for the patient, not the doctor. So we can understand that this book is focused on the practical tips at an actual emergency situation. 2. "GuGeupGanIBang(救急簡易方)" is the most complete emergency treatment book and a Hangul medical book, which is scarce. It was published shortly after the invention of Hangul. It is an important material showing how terms related to medicine such as disease, prescriptions, symptoms were understood and expressed. As it is important in both medical book presswork study and Korean language study, it is designated as treasure. 3. The existing "GuGeupGanIBang(救急簡易方)" seem to be a reissued version. It is also not complete and scattered in several places. Volumes 1,2,3,6 and 7 survived, and volumes 4,5 and 8 have not been found yet.

• Journal of Broadcast Engineering
• /
• v.25 no.2
• /
• pp.143-150
• /
• 2020

Huge amounts of contents are being uploaded every day on various streaming platforms. Among those videos, game and sports videos account for a great portion. The broadcasting companies sometimes create and provide highlight videos. However, these tasks are time-consuming and costly. In this paper, we propose models that automatically predict highlights in games and sports matches. While most previous approaches use visual information exclusively, our models use both audio and visual information, and present a way to understand short term and long term flows of videos. We also describe models that combine GAN to find better highlight features. The proposed models are evaluated on e-sports and baseball videos.