• Proceedings of the Korean Society of Computer Information Conference
• /
• 2024.01a
• /
• pp.45-46
• /
• 2024

본 논문에서는 Self-Attention 기반 딥러닝 기법인 Dense Prediction Transformer(DPT) 모델을 MVTec Anomaly Detection(MVTec AD) 데이터셋에 적용하여 실제 산업 제품 이미지 내 이상 부분을 분할하는 연구를 진행하였다. DPT 모델의 적용을 통해 기존 Convolutional Neural Network(CNN) 기반 이상 탐지기법의 한계점인 지역적 Feature 추출 및 고정된 수용영역으로 인한 문제를 개선하였으며, 실제 산업 제품 데이터에서의 이상 분할 시 기존 주력 기법인 U-Net의 구조를 적용한 최고 성능의 모델보다 1.14%만큼의 성능 향상을 보임에 따라 Self-Attention 기반 딥러닝 기법의 적용이 산업 제품 이상 분할에 효과적임을 입증하였다.

• The Journal of the Acoustical Society of Korea
• /
• v.40 no.2
• /
• pp.169-175
• /
• 2021

Recently, an ultrasound hologram and its applications have gained attention in the ultrasound research field. However, the determination technique of transmit signal phases, which generate a hologram, has not been significantly advanced from the previous algorithms which are time-consuming iterative methods. Thus, we applied the deep learning technique, which has been previously adopted to generate an optical hologram, to generate an ultrasound hologram. We further examined the Deep learning-based Holographic Ultrasound Generation algorithm (Deep-HUG). We implement the U-Net-based algorithm and examine its generalizability by training on a dataset, which consists of randomly distributed disks, and testing on the alphabets (A-Z). Furthermore, we compare the Deep-HUG with the previous algorithm in terms of computation time, accuracy, and uniformity. It was found that the accuracy and uniformity of the Deep-HUG are somewhat lower than those of the previous algorithm whereas the computation time is 190 times faster than that of the previous algorithm, demonstrating that Deep-HUG has potential as a useful technique to rapidly generate an ultrasound hologram for various applications.

• Nuclear Engineering and Technology
• /
• v.55 no.6
• /
• pp.2276-2282
• /
• 2023

We deploy artificial neural networks to unfold neutron spectra from measured energy-integrated quantities. These neutron spectra represent an important parameter allowing to compute the absorbed dose and the kerma to serve radiation protection in addition to nuclear safety. The built architectures are inspired from convolutional neural networks. The first architecture is made up of residual transposed convolution's blocks while the second is a modified version of the U-net architecture. A large and balanced dataset is simulated following "realistic" physical constraints to train the architectures in an efficient way. Results show a high accuracy prediction of neutron spectra ranging from thermal up to fast spectrum. The dataset processing, the attention paid to performances' metrics and the hyper-optimization are behind the architectures' robustness.

• Journal of Broadcast Engineering
• /
• v.27 no.1
• /
• pp.69-79
• /
• 2022

In this study, a method of generating and utilizing a normal map image used to represent the characteristics of the surface of an image material to improve the classification accuracy of the original material image is proposed. First of all, (1) to generate a normal map that reflects the surface properties of a material in an image, a U-Net with attention-R2 gate as a generator was used, and a Pix2Pix-based method using the generated normal map and the similarity with the original normal map as a reconstruction loss was used. Next, (2) we propose a network that can improve the accuracy of classification of the original material image by applying the previously created normal map image to the attention gate of the classification network. For normal maps generated using Pixar Dataset, the similarity between normal maps corresponding to ground truth is evaluated. In this case, the results of reconstruction loss function applied differently according to the similarity metrics are compared. In addition, for evaluation of material image classification, it was confirmed that the proposed method based on MINC-2500 and FMD datasets and comparative experiments in previous studies could be more accurately distinguished. The method proposed in this paper is expected to be the basis for various image processing and network construction that can identify substances within an image.

• Spatial Information Research
• /
• v.22 no.6
• /
• pp.45-54
• /
• 2014

Establishment of safety net for the socially disadvantaged attracts large attention because of the recent crime increasing against vulnerable groups. For the successful establishment of social safety net, the test-bed for evaluation and realization of crime-related research results is required. However, previous R&D test-bed projects such as The Korean Land Specialization Program or U-Eco City project remains only to the stage of verification. Therefore, there are limitedness for realization of result technologies or sustainable operation & management of test-bed after projects finished. So, sustainable operation & management system and guideline of test-bed are necessary. Therefore, this study reviews the strengths and weaknesses of existing test-bed cases and intelligent security researches. After reviewing, the concept of a Intelligent Security Test-bed is established and appropriate test-bed selection criteria is also suggested. Based on objective criteria, selected test-bed can achieve sustainable management even after finishing the project and contribute the construction of standard model for citizen's safety.

• Nuclear Engineering and Technology
• /
• v.53 no.8
• /
• pp.2629-2639
• /
• 2021

Within the frame of the EMPIrE test, four monolithic mini-plates were irradiated in the ATR reactor. In two of them, the monolithic U(Mo) foil had been PVD-coated with Zr before the plate manufacturing. Extensive microstructural characterizations were performed on a fresh archive mini-plate, using Optical Microscopy (OM), Scanning Electron Microscopy (SEM) combined with Energy Dispersive Spectroscopy (EDS), Electron Backscattered Diffraction (EBSD) and Focused Ion Beam (FIB)/Transmission Electron Microscopy (TEM) with nano EDS. A particular attention was paid to the examination of the U(Mo) foil, the PVD coating, the cladding/Zr and Zr/U(Mo) interfaces. The Zr coating has a thickness around 15 ㎛. It has a columnar microstructure and appears dense. The cohesion of the cladding/Zr and Zr/U(Mo) interfaces seems to be satisfactory. An almost continuous layer with a thickness of the order of 100-300 nm is present at the cladding/Zr interface and corresponds to an oxidized part of the Zr coating. At the Zr/U(Mo) interface, a thin discontinuous layer is observed. It could correspond to locally oxidized U(Mo). This work provides a basis for interpreting the results of characterizations on EMPIrE irradiated plates.

• Nuclear Engineering and Technology
• /
• v.54 no.10
• /
• pp.3631-3640
• /
• 2022

Uranium mill tailing ponds (UMTPs) are risk source of debris flow and a critical source of environmental U and Rn pollution. The technology of microbial induced calcium carbonate precipitation (MICP) has been extensively studied on reinforcement of UMTs, while little attention has been paid to the effects of MICP on U & Rn release, especially when incorporation of metakaolin and bacillus subtilis (MBS). In this study, the reinforcement and U & Rn immobilization role of MBS -MICP solidification in different grouting cycle for uranium mill tailings (UMTs) was comprehensively investigated. The results showed that under the action of about 166.7 g/L metakaolin and ~50% bacillus subtilis, the solidification cycle of MICP was shortened by 50%, the solidified bodies became brittle, and the axial stress increased by up to 7.9%, and U immobilization rates and Rn exhalation rates decrease by 12.6% and 0.8%, respectively. Therefore, the incorporation of MBS can enhance the triaxial compressive strength and improve the immobilization capacity of U and Rn of the UMTs bodies solidified during MICP, due to the reduction of pore volume and surface area, the formation of more crystals general gypsum and gismondine, as well as the enhancing of coprecipitation and encapsulation capacity.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.1
• /
• pp.60-79
• /
• 2022

The U-Net architecture-based segmentation models attained remarkable performance in numerous medical image segmentation missions like skin lesion segmentation. Nevertheless, the resolution gradually decreases and the loss of spatial information increases with deeper network. The fusion of adjacent layers is not enough to make up for the lost spatial information, thus resulting in errors of segmentation boundary so as to decline the accuracy of segmentation. To tackle the issue, we propose a new deep learning-based segmentation model. In the decoding stage, the feature channels of each decoding unit are concatenated with all the feature channels of the upper coding unit. Which is done in order to ensure the segmentation effect by integrating spatial and semantic information, and promotes the robustness and generalization of our model by combining the atrous spatial pyramid pooling (ASPP) module and channel attention module (CAM). Extensive experiments on ISIC2016 and ISIC2017 common datasets proved that our model implements well and outperforms compared segmentation models for skin lesion segmentation.

• Journal of Korea Multimedia Society
• /
• v.23 no.2
• /
• pp.186-203
• /
• 2020

Image-based Virtual Try-On Systems are among the most potential solution for virtual fitting which tries on a target clothes into a model person image and thus have attracted considerable research efforts. In many cases, current solutions for those fails in achieving naturally looking virtual fitted image where a target clothes is transferred into the body area of a model person of any shape and pose while keeping clothes context like texture, text, logo without distortion and artifacts. In this paper, we propose a new improved image-based virtual try-on network system based on keypoints, which we name as KP-VTON. The proposed KP-VTON first detects keypoints in the target clothes and reliably predicts keypoints in the clothes of a model person image by utilizing a dense human pose estimation. Then, through TPS transformation calculated by utilizing the keypoints as control points, the warped target clothes image, which is matched into the body area for wearing the target clothes, is obtained. Finally, a new try-on module adopting Attention U-Net is applied to handle more detailed synthesis of virtual fitted image. Extensive experiments on a well-known dataset show that the proposed KP-VTON performs better the state-of-the-art virtual try-on systems.

• Ocean Science Journal
• /
• v.40 no.3
• /
• pp.177-182
• /
• 2005

The process of embryogenesis and larval development of the asteroid sea star Asterias amurensis $(U{\ddot{u}}tken)$ was observed, with special attention paid to morphological change and larval duration. In reproductive season, mature sea stars were collected under floating net cages, located in Tongyeong, southern Korea. The mature eggs are $138\;{\mu}m$ in average diameter, semi-translucent and orange in color, sperms in good condition appear light cream to white-gray in color. Embryos develop through the holoblastic equal cleavage stage and a wrinkled blastula stage that lasts about 9 hours after fertilization. Gastrulae bearing an expanded archenteron hatch from the fertilization envelope 22 hours after fertilization. At the end of gastrulation, rudiments of the left and right coelom are formed. By day 2, larvae possess complete alimentary canal and begin to feed. At this stage, the larva is called early bipinnaria. In 6-day-old larvae, the pre- and post- oral ciliated bands form complete circuits and the bipinnarial processes start to develop. By day 12, the lateral and anterior projection of the larval wall processes along the ciliated bands begins to thicken and curl, and the ciliated bands become more prominent. By day 32, early brachiolaria are presented with three pairs of brachiolar arms. Advanced brachiolaria with a well-developed brachiolar complex (three pairs of brachia and central adhesive disc) occur 6 weeks after fertilization. In the field, spawning of the sea star was observed in April to May, settlement form larvae and just settlements seem to occur from June to July, and early juveniles occur from August to September. Although we had not described the end of brachiolaria stage, it can be tentatively estimated that the duration of the pelagic stage of A. amurensis is 40 to 50 days.