• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권8호
• /
• pp.3942-3961
• /
• 2019

Recently, the models of deep super-resolution networks can successfully learn the non-linear mapping from the low-resolution inputs to high-resolution outputs. However, for large scaling factors, this approach has difficulties in learning the relation of low-resolution to high-resolution images, which lead to the poor restoration. In this paper, we propose Stage Generative Adversarial Networks (Stage-GAN) with semantic maps for image super-resolution (SR) in large scaling factors. We decompose the task of image super-resolution into a novel semantic map based reconstruction and refinement process. In the initial stage, the semantic maps based on the given low-resolution images can be generated by Stage-0 GAN. In the next stage, the generated semantic maps from Stage-0 and corresponding low-resolution images can be used to yield high-resolution images by Stage-1 GAN. In order to remove the reconstruction artifacts and blurs for high-resolution images, Stage-2 GAN based post-processing module is proposed in the last stage, which can reconstruct high-resolution images with photo-realistic details. Extensive experiments and comparisons with other SR methods demonstrate that our proposed method can restore photo-realistic images with visual improvements. For scale factor ${\times}8$, our method performs favorably against other methods in terms of gradients similarity.

• 방송공학회논문지
• /
• 제27권3호
• /
• pp.437-450
• /
• 2022

생성적 적대 신경망(GAN, Generative Adversarial Networks)는 이미지 생성 분야에서 주목할 만한 발전을 이루었다. 하지만 큰 데이터 셋에서 불안정한 모습을 보인다는 한계 때문에 다양한 응용 분야에 쉽게 적용하기 어렵다. 단일 이미지 생성적 적대 신경망은 한장의 이미지의 내부 분포를 잘 학습하여 다양한 영상을 생성하는 분야이다. 큰 데이터셋이 아닌 단 한장만 학습함으로써 안정적인 학습이 가능하며 이미지 리타겟팅, 이미지 조작, super resolution 등 다양한 분야에 활용 가능하다. 본 논문에서는 SinGAN, ConSinGAN, InGAN, DeepSIM, 그리고 One-Shot GAN 총 다섯 개의 단일 이미지 생성적 적대 신경망을 살펴본다. 우리는 각각의 단일 이미지 생성적 적대 신경망 모델들의 성능을 비교하고 장단점을 분석한다.

• International Journal of Contents
• /
• 제16권2호
• /
• 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.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권7호
• /
• pp.2547-2567
• /
• 2021

It has been widely acknowledged that occlusion impairments adversely distress many face recognition algorithms' performance. Therefore, it is crucial to solving the problem of face image occlusion in face recognition. To solve the image occlusion problem in face recognition, this paper aims to automatically de-occlude the human face majority or discriminative regions to improve face recognition performance. To achieve this, we decompose the generative process into two key stages and employ a separate generative adversarial network (GAN)-based network in both stages. The first stage generates an initial coarse face image without an occlusion mask. The second stage refines the result from the first stage by forcing it closer to real face images or ground truth. To increase the performance and minimize the artifacts in the generated result, a new refine loss (e.g., reconstruction loss, perceptual loss, and adversarial loss) is used to determine all differences between the generated de-occluded face image and ground truth. Furthermore, we build occluded face images and corresponding occlusion-free face images dataset. We trained our model on this new dataset and later tested it on real-world face images. The experiment results (qualitative and quantitative) and the comparative study confirm the robustness and effectiveness of the proposed work in removing challenging occlusion masks with various structures, sizes, shapes, types, and positions.

• Advances in Traditional Medicine
• /
• 제5권2호
• /
• pp.75-91
• /
• 2005

Great efforts have been made to improve both the quality of life and life expectancy of diabetes by treating problems associated with chronic complications such as neuropathy, retinopathy and nephropathy. In particular, diabetes is an increased risk of developing several types of kidney disease, and the predominant cause of end-stage renal disease in patients with this disorder is diabetic nephropathy. Therefore, prevention of the occurrence and progression of diabetes and its complications has become a very important issue. The scientific observations of an animal model of streptozotocin-induced diabetes, spontaneously occurring diabetes and diabetic nephropathy in this study suggest that one of the Kampo prescriptions, Hachimi-jio-gan comprising eight constituents, is a novel therapeutic agent.

• 전자통신동향분석
• /
• 제35권4호
• /
• pp.91-102
• /
• 2020

Recently, generative adversarial networks (GANs) is a field of research that has rapidly emerged wherein many studies conducted shows overwhelming results. Initially, this was at the level of imitating the training dataset. However, the GAN is currently useful in many fields, such as transformation of data categories, restoration of erased parts of images, copying facial expressions of humans, and creation of artworks depicting a dead painter's style. Although many outstanding research achievements have been attracting attention recently, GANs have encountered many challenges. First, they require a large memory facility for research. Second, there are still technical limitations in processing high-resolution images over 4K. Third, many GAN learning methods have a problem of instability in the training stage. However, recent research results show images that are difficult to distinguish whether they are real or fake, even with the naked eye, and the resolution of 4K and above is being developed. With the increase in image quality and resolution, many applications in the field of design and image and video editing are now available, including those that draw a photorealistic image as a simple sketch or easily modify unnecessary parts of an image or a video. In this paper, we discuss how GANs started, including the base architecture and latest technologies of GANs used in high-resolution, high-quality image creation, image and video editing, style translation, content transfer, and technology.

• 한국콘텐츠학회논문지
• /
• 제21권12호
• /
• pp.322-330
• /
• 2021

수산 양식장에서 어류 질병을 초기에 발견하지 못하는 경우 밀폐된 공간 안에서 확산하기 때문에 집단 폐사로 이어질 확률이 매우 높다. 이런 이유로 질병의 조기 발견은 양식업에서 매우 중요하다. 양식장에서 질병의 확산을 막기 위해서는 초기에 병이 든 어류를 자동식별이 가능한 방법이 필요하다. 최근 딥러닝 기반의 어류 질병 자동식별 방법이 많이 사용되고 있는데, 어류의 질병 이미지가 충분하지 않아 객체 식별에 많은 어려움이 있다. 본 논문은 질병 자동식별 예측을 위한 질병 이미지의 부족 문제를 해결하기 위해서 SinGAN 딥러닝 모델을 이용하여 정상 이미지와 질병 이미지를 합성해 다양한 어류 질병 이미지를 자동 생성하는 방법을 제안한다. 넙치에서 가장 빈번히 발생하는 3가지 질병 스쿠티카병, 비브리오증, 림포시스티스에 대해서 SinGAN 기반으로 질병 이미지를 증강한다. 본 연구에서는 넙치 정상 이미지 11장에 각 질병 패턴 10가지를 합성하여서 스쿠티카병 110장, 비브리오증 110장, 림포시스티스 110장으로 총 330장을 만들었고 이를 통해 생성된 이미지는 4배수 하여 1,320장의 이미지를 생성할 수 있었다.

• 대한원격탐사학회지
• /
• 제38권6_1호
• /
• pp.1357-1369
• /
• 2022

구름 제거는 식생 모니터링, 변화 탐지 등과 같은 광학 영상이 필요한 모든 작업에서 필수적인 영상 처리 과정이다. 이 논문에서는 조건부 생성적 적대 신경망(conditional generative adversarial networks, cGANs)과 회귀 기반 보정을 결합하여 구름이 없는 시계열 광학 영상 세트를 구성하는 2단계의 구름 제거 기법을 제안하였다. 첫 번째 단계에서는 광학 영상과 synthetic aperture radar 영상 간 정량적 관계를 이용하는 cGANs을 이용하여 초기 예측 결과를 생성한다. 두 번째 단계에서는 구름이 아닌 영역에서 예측 결과와 실제 값과의 관계를 random forest 기반 회귀 모델링을 통해 정량화한 후에 cGANs 기반 예측 결과를 보정한다. 제안 기법은 김제의 벼 재배지에서 Sentinel-2 영상과 COSMO-SkyMed 영상을 이용한 구름 제거 실험을 통해 적용 가능성을 평가하였다. cGAN 모델은 구름 영역에서 지표면 상태의 급격한 변화가 발생하는 논 재배지를 대상으로 반사율 값을 효과적으로 예측할 수 있었다. 또한 두 번째 단계의 회귀 기반 보정은 예측 대상 영상에서 시간적으로 떨어진 보조 영상을 이용하는 회귀 기반 구름 제거 기법에 비해 예측 정확도를 향상시킬 수 있었다. 이러한 실험 결과는 구름이 없는 광학 영상을 환경 모니터링에 이용할 수 없는 경우 제안된 방법이 구름 오염 지역을 복원하는데 효과적으로 적용될 수 있음을 나타낸다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권3호
• /
• pp.881-895
• /
• 2023

Modern image inpainting techniques based on deep learning have achieved remarkable performance, and more and more people are working on repairing more complex and larger missing areas, although this is still challenging, especially for facial image inpainting. For a face image with a huge missing area, there are very few valid pixels available; however, people have an ability to imagine the complete picture in their mind according to their subjective will. It is important to simulate this capability while maintaining the identity features of the face as much as possible. To achieve this goal, we propose a three-stage network model, which we refer to as the identity and structure feature refinement network (ISFRNet). ISFRNet is based on 1) a pre-trained pSp-styleGAN model that generates an extremely realistic face image with rich structural features; 2) a shallow structured network with a small receptive field; and 3) a modified U-net with two encoders and a decoder, which has a large receptive field. We choose structural similarity index (SSIM), peak signal-to-noise ratio (PSNR), L1 Loss and learned perceptual image patch similarity (LPIPS) to evaluate our model. When the missing region is 20%-40%, the above four metric scores of our model are 28.12, 0.942, 0.015 and 0.090, respectively. When the lost area is between 40% and 60%, the metric scores are 23.31, 0.840, 0.053 and 0.177, respectively. Our inpainting network not only guarantees excellent face identity feature recovery but also exhibits state-of-the-art performance compared to other multi-stage refinement models.

• 한국BIM학회 논문집
• /
• 제11권2호
• /
• pp.54-59
• /
• 2021

In the process of designing a building, the process of expressing the designer's ideas through images is essential. However, it is expensive and time consuming for a designer to analyze every individual case image to generate a hypothetical design. This study aims to visualize the basic design draft sketch made by the designer as a real image using the Generative Adversarial Network (GAN) based on the continuously accumulated architectural case images. Through this, we proposed a method to build an automated visualization environment using artificial intelligence and to visualize the architectural idea conceived by the designer in the architectural planning stage faster and cheaper than in the past. This study was conducted using approximately 20,000 images. In our study, the GAN algorithm allowed us to represent primary materials and shades within 2 seconds, but lacked accuracy in material and shading representation. We plan to add image data in the future to address this in a follow-up study.