• Journal of Digital Convergence
• /
• v.20 no.1
• /
• pp.265-275
• /
• 2022

This study analyzes and categorizes the cases of military look's transitional design ideas in recent women's fashion collections, and derives characteristics. The research method is a theoretical review of military look and an analysis of fashion collection cases. The research results were classified into a total of six transformational design ideas. As a structural change in design, it is a decentralized type, a type of expansion and reduction, a change in the entire material, or a transition of some materials, and finally a type according to heterogeneous harmony and organic combination corresponding to styling. Finally, a total of three characteristics are the reconstruction of structural elements, the expansion of the metric of the second mix match, and the emotional fusion of styling. I hope that the study of the transformative type of idea of the new military look will be the driving force for creative design development and will be a basic study that can read the current status and changes of the times throughout fashion design.

• Korean Journal of Cognitive Science
• /
• v.34 no.3
• /
• pp.197-226
• /
• 2023

Generative Text Summarization is one of the Natural Language Processing tasks. It generates a short abbreviated summary while preserving the content of the long text. ROUGE is a widely used lexical-overlap based metric for text summarization models in generative summarization benchmarks. Although it shows very high performance, the studies report that 30% of the generated summary and the text are still inconsistent. This paper proposes a methodology for evaluating the performance of the summary model without using the correct summary. AggreFACT is a human-annotated dataset that classifies the types of errors in neural text summarization models. Among all the test candidates, the two cases, generation summary, and when errors occurred throughout the summary showed the highest correlation results. We observed that the proposed evaluation score showed a high correlation with models finetuned with BART and PEGASUS, which is pretrained with a large-scale Transformer structure.

• The Research Journal of the Costume Culture
• /
• v.32 no.4
• /
• pp.577-596
• /
• 2024

Geometric abstract art has been a persistent form since ancient times and is defined in modernity as a genre of abstract painting. According to previous studies, the geo- metric structure of traditional Korean costume possesses uniqueness based on Korean thought and philosophy. Therefore, the purpose of this study is to first derive the characteristics of Wilhelm Worringer's theory, a theorist of geometric abstraction, and then compare the spiritual similarities between the traditional Korean costume and geometric abstract art, with the aim of examining the artistic competitiveness of Hanbok. Photographic materials from domestic and international museums, as well as books published by museums and government agencies, were analyzed. Further, the characteristics of Worringer's theory of the abstract impulse were derived. The results are as follows. First, the external factors are characterized by a single geometric shape on a flat surface with restrained spatial expression. The internal factors include anxiety, comfort from religion or ideology, and the need for a visual refuge for rest. Second, traditional Korean costume and geometric abstraction share common spiritual contexts of artistry, transcendence, and stability according to Worringer's theory. Third, the study of the spiritual similarities between the traditional Korean costume and geometric abstract art through Worringer's theory reveals that the artistry of traditional Korean costume is expressed in universal forms, transcendence as intrinsic origins, and stability as a unique aesthetic consciousness. These findings provide a framework for reinterpreting the originality and universality of traditional Korean costume as Korean cultural content, infused with the characteristics of modern art.

• Structural Engineering and Mechanics
• /
• v.91 no.5
• /
• pp.443-457
• /
• 2024

At present, the traditional concrete surface inspection methods based on artificial vision have the problems of high cost and insecurity, while the computer vision methods rely on artificial selection features in the case of sensitive environmental changes and difficult promotion. In order to solve these problems, this paper introduces deep learning technology in the field of computer vision to achieve automatic feature extraction of structural damage, with excellent detection speed and strong generalization ability. The main contents of this study are as follows: (1) A method based on DeepLabV3+ convolutional neural network model is proposed for surface detection of post-earthquake structural damage, including surface damage such as concrete cracks, spaling and exposed steel bars. The key semantic information is extracted by different backbone networks, and the data sets containing various surface damage are trained, tested and evaluated. The intersection ratios of 54.4%, 44.2%, and 89.9% in the test set demonstrate the network's capability to accurately identify different types of structural surface damages in pixel-level segmentation, highlighting its effectiveness in varied testing scenarios. (2) A semantic segmentation model based on DeepLabV3+ convolutional neural network is proposed for the detection and evaluation of post-earthquake structural components. Using a dataset that includes building structural components and their damage degrees for training, testing, and evaluation, semantic segmentation detection accuracies were recorded at 98.5% and 56.9%. To provide a comprehensive assessment that considers both false positives and false negatives, the Mean Intersection over Union (Mean IoU) was employed as the primary evaluation metric. This choice ensures that the network's performance in detecting and evaluating pixel-level damage in post-earthquake structural components is evaluated uniformly across all experiments. By incorporating deep learning technology, this study not only offers an innovative solution for accurately identifying post-earthquake damage in civil engineering structures but also contributes significantly to empirical research in automated detection and evaluation within the field of structural health monitoring.

• Journal of the Korean Society of Costume
• /
• v.30
• /
• pp.69-84
• /
• 1996

The main current of ancient Korean thought is based on Yuk(易) : which solves the principle of creation of universe letter(文字) which is the principle of geometry and three elements thought of circle square . It's related to the creation principle of the universe and regarded as a sign that demons-trates shows the way to read the mathematical principle. The original form of Korean clothes(Han-Bock, 韓服) is represents the structure of hu-man body which reflects a small universe. So a good structure and meaning of the human body is well read in the formation of Korean clothes(Han-Bock, 韓服) and that is a good symbolization of the clothes(Han-Bock, 韓服) as the space of small universe. The good formation of Korean clothes(Han-Bock, 韓服) could be shown by cutting out straight bending twisting and turning each straitly cutting piece is applied to its trans-tormation. Geametical formation with obvious sym-metric dividing of front-back left-right and top-bottom is well shown in Korean clothes by twisting Yuk(易) or m bius strips. So we can find out whole formation by representing only either 'front and back' or 'left and right' and also sa-pok(사폭) of man's trousers seop(섶) of jeo-go-ri(short jacket 저고리) kyut-ma-gi(곁마기) of Sam-hwai-jang jeo-go-ri (삼회장 저고리), mu(무) of jig-ryung(straight collar and long and wide sleeved robe, 直領) trouser wide and ga-rae-ba-dae(가래바대) of dan-cok-kok(woman's under trousers). The formation line of Korean clothes(Han-Bock, 韓服) is based on theprinciple of unlimited m bius strips by twisting of turning direction from universal principle and original basic form is not changd even by turning in-side out. Unity of the whole and part in Korean clothes(Han-Bock, 韓服) could be found in nonorientable thought(非始原思想) which represents the unity and dependence kil(길) of jeo-go-ri(short jacket, 저고리) and po(long and wide sleeved robe, 袍). Selva-gewise(식서방향) of sleeves and seam of out-side of kil(길) and outside of seop(섶) are directed to outside of seop(섶) inside of seop(섶) is to kil and direction of inseam of kheut-dong(끝동) and kyut-ma-gi(곁마기) is directed to sleeves. Korean clothes(Han-Bock, 韓服) is usually made by cutting fabric for several parts and sewing them and the way to make Korean clothes is deeply related to the theory of chon-pu-kyung jong-il and il-seok -sam-geuk. As the development of men's consciousness is pro-gressed from total separation to separation and superseparation. Sewing process of korean clothes(Han-Bock, 韓服) can be make single dimension to double or triple dimension of korean clothes(Han-Bock, 韓服) silhoutte bacause it can be include principle of topology as noneuclity.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.40 no.5
• /
• pp.397-409
• /
• 2003

This paper describes an implementation of radix-4 trellis parallel architecture and backward state transition control trace back Viterbi decoder, and presents the application results to high speed wireless LAN. The radix-4 parallelized architecture Vietrbi decoder can not only improve the throughput with simple structure, but also have small processing delay time and overhead circuit compared to M-step trellis architecture one. Based on these features, this paper addresses a novel Viterbi decoder which is composed of branch metric computation, architecture of ACS and trace back decoding by sequential control of backward state transition for the implementation of radix-4 trellis parallelized structure. With the proposed architecture, the decoding of variable code rate due to puncturing the base code can easily be implemented by the unified Viterbi decoder. Moreover, any additional circuit and/or peripheral control logic are not required in the proposed decoder architecture. The trace back decoding scheme with backward state transition control can carry out the sequential decoding according to ACS cycle clock without additional circuit for survivor memory control. In order to evaluate the usefulness, the proposed method is applied to channel CODEC of the IEEE 802.11a high speed wireless LAN, and HDL coding simulation results are presented.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.15 no.4
• /
• pp.81-95
• /
• 2012

This study was intended for Fagus engleriana stand in Is. Ulleung where the disturbance of vegetation has been caused by the exploitation and the increase of tourists. For the effective conservation and management on this issue, this study was conducted provide basic data. The sixteen study sites ($20{\times}20m$) were installed in the dominant Fagus engleriana stand and the base environment and vegetation were investigated. The Fagus engleriana stand was classified into two groups, The Fagus engleriana stand was classified into two groups, community A is Fagus engleriana-Sorbus amurensis and community B is Fagus engleriana-Acer pictum subsp. Mono by cluster analysis and community A were nothing signigicant by indicator species analysis. Community B were Eight species (Tsuga sieboldii, Camellia japonica, Dystaenia takesimana ect.) significant by indicator species analysis. The diameter class of 16cm to 25cm was 53.7% in population structure of Fagus engleriana, which was the highest and showed inverse J-distribution. Species diversity index (H') of investigated woody layer group ranged from 0.99 to 2.05 and that of under layer group ranged from 1.75~2.59. According to Non-metric Multidimensional Scaling (NMS) analysis, the woody layer was divided into community A developed in the region having relatively high sand content at high altitudes and community B formed at the place having relatively high clay content at low altitudes. Then this classification was significant through Multi-Response Permutation Procedures (MRPP) analysis. The distribution of understory vegetation through Detrended Correspondence Analysis (DCA) was induced by the silt content and cover degree of vegetation layer.

• Economic and Environmental Geology
• /
• v.29 no.3
• /
• pp.315-323
• /
• 1996

An apatite drain was constructed on September 30, 1994 at the Green Valley Abandoned Coal Mine site near Terre Haute in west central Indiana. The primary objective of this experiment is to evaluate the long-term ability of the apatite drain to mitigate acid mine drainage (AMD) under field conditions. The drain 9 m long, 3.3 m wide, and 0.75 m deep, contain 95 rum to No. 30 mesh-size apatite ore (francolite) and receive AMD seepage from reclaimed gob piles, and designed according to the laboratory testing. The apatite drain was covered with limestone riprap and filter fabric to protect the drainage system from stormwater and siltation. The drain consists of about 50 metric tons of apatite ore obtained from a phosphate mine in Florida. A gabion structure was constructed downstream of the apatite drain to create a settling pond to collect precipitates. Apatite effectively removed iron up to 4,200 mg/l, aluminum up to 830 mg/l and sulfate up to 13,430 mg/l. The pH was nearly constant for the influent and effluent, ranging between 3.1 and 4.3. Flow rate measured at the gabion structure ranged from 3 to 4.5 l/m. Precipitates of iron and aluminum phosphate (yellow and white suspendid solids) continued to accumulate in the settling pond.

• The Sea:JOURNAL OF THE KOREAN SOCIETY OF OCEANOGRAPHY
• /
• v.26 no.3
• /
• pp.167-182
• /
• 2021

This study was carried out to investigate temporal and spatial distribution of macrobenthic community and elucidate effects of environmental factors on change of community structure in an intertidal flat, Hakseong-ri, Boryeong, Korea. Field surveys were seasonally conducted to collect samples of sediment and macrobenthos using can core in triplicate at nine stations in 2016 and 2017. Our results showed that sediment had high mud content (above 60%) in most samples and mean content of loss on ignition was 2.3% in 2016. A total of 79 species was collected in the study site during the study period. Mean density and biomass were 611 ind./m² and 64.1 gWWt/m², respectively. Heteromastus filiformis was the dominant species (48.6%, 297 ind./m²) followed by Macrophthalmus japonicus (10.1%, 62 ind./m²) and Upogebia major (6.9%, 42 ind./m²). Three assembly groups resulted from cluster analysis were more distinguished by interaction between organisms and frequency of dominant species than by physical and chemical environment characteristics. In addition, macrobenthic community in the Hakseong intertidal flat showed seasonal changes based on non metric multidimensional scaling using species abundance.

• Korean Journal of Remote Sensing
• /
• v.39 no.2
• /
• pp.143-156
• /
• 2023

To precisely and reliably analyze the contents of the satellite imagery, recognizing the clouds which are the obstacle to gathering the useful information is essential. In recent times, deep learning yielded satisfactory results in various tasks, so many studies using deep neural networks have been conducted to improve the performance of cloud detection. However, existing methods for cloud detection have the limitation on increasing the performance due to the adopting the network models for semantic image segmentation without modification. To tackle this problem, we introduced the multi-branch neural architecture search to find optimal network structure for cloud detection. Additionally, the proposed method adopts the soft intersection over union (IoU) as loss function to mitigate the disagreement between the loss function and the evaluation metric and uses the various data augmentation methods. The experiments are conducted using the cloud detection dataset acquired by Arirang-3/3A satellite imagery. The experimental results showed that the proposed network which are searched network architecture using cloud dataset is 4% higher than the existing network model which are searched network structure using urban street scenes with regard to the IoU. Also, the experimental results showed that the soft IoU exhibits the best performance on cloud detection among the various loss functions. When comparing the proposed method with the state-of-the-art (SOTA) models in the field of semantic segmentation, the proposed method showed better performance than the SOTA models with regard to the mean IoU and overall accuracy.