• 한국HCI학회:학술대회논문집
• /
• 2008.02a
• /
• pp.19-26
• /
• 2008

This paper is related to a new technology allowing a user to have a 360 degree viewpoint of the virtual wearing object, and to an object VR(Virtual Reality)-based 2D virtual textile wearing system using viewpoint vector estimation and textile texture mapping. The proposed system is characterized as capable of virtually wearing a new textile pattern selected by the user to the clothing shape section segmented from multiview 2D images of clothes model for object VR, and three-dimensionally viewing its virtual wearing appearance at a 360 degree viewpoint of the object. Regardless of color or intensity of model clothes, the proposed system is possible to virtually change the textile pattern with holding the illumination and shading properties of the selected clothing shape section, and also to quickly and easily simulate, compare, and select multiple textile pattern combinations for individual styles or entire outfits. The proposed system can provide higher practicality and easy-to-use interface, as it makes real-time processing possible in various digital environment, and creates comparatively natural and realistic virtual wearing styles, and also makes semi -automatic processing possible to reduce the manual works to a minimum. According to the proposed system, it can motivate the creative activity of the designers with simulation results on the effect of textile pattern design on the appearance of clothes without manufacturing physical clothes and, as it can help the purchasers for decision-making with them, promote B2B or B2C e-commerce.

• The Journal of the Korea Contents Association
• /
• v.11 no.6
• /
• pp.106-115
• /
• 2011

In this paper, we evaluate the effectiveness of previous shading algorithms in depicting shape of 3d objects. We perform a study in which people are shown an image of one of ten 3D objects shaded with one of eight styles and asked to orient a gauge to coincide with the surface normal at many positions on the object's surface. The normal estimates are compared with each other and with ground truth data provided by a registered 3D surface model to analyze accuracy and precision. Our experiments suggest that people interpret certain shape differently depending on shading of 3d object. This paper offers substantial evidence that current computer graphics shading algorithms can effectively depict shape of 3d objects where the algorithms have the properties of lots of tone steps and uniformly distributed tone steps. This type of analysis can guide the future development of new CG shading algorithms in computer graphics for the purpose of shape perception.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.4
• /
• pp.457-461
• /
• 2010

Even though face recognition researches have been developed for a long ago, there is no practical face recognition system in real life. It is caused by several real situations where non-facial components such as glasses, scarf, and hair occlude facial components while facial images in a face database are well designed. This occlusion decreases recognition performance. Previous approaches in recent years have tried to solve non-facial components but have not resulted in enough performance. In this paper, we propose a method to handle this problem based on support vector data description, which trains the hyperball in feature space to find the minimum distance estimating the approximated face. In order to evaluate its performance and validate the effectiveness of the proposed method, we make several experiments and the results show that the proposed method has a considerable effectiveness.

• Journal of Internet Computing and Services
• /
• v.19 no.4
• /
• pp.27-33
• /
• 2018

In the process of submitting a manuscript to a journal in order to present the results of the research at the research institution, researchers often proofread the manuscript because it can manuscripts to communicate the results more effectively. Currently, most of the manuscript proofreading companies use the manual proofreader assignment method according to the subjective judgment of the matching manager. Therefore, in this paper, we propose a topic-based proofreader matching method for effective proofreading results. The proposed method is categorized into two steps. First, a topic modeling is performed by using Latent Dirichlet Allocation. In this process, the frequency of each document constituting the representative document of a user is determined according to the importance of the document. Second, the user similarity is calculated based on the cosine similarity method. In addition, we confirmed through experiments by using real-world dataset. The performance of the proposed method is superior to the comparative method, and the validity of the matching results was verified using qualitative evaluation.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.22 no.3
• /
• pp.1-8
• /
• 2022

Recently, Google MedeiaPipe presents a novel approach for neural network-based hair segmentation from a single camera input specifically designed for real-time, mobile application. Though neural network related to hair segmentation is relatively small size, it produces a high-quality hair segmentation mask that is well suited for AR effects such as a realistic hair recoloring. However, it has undesirable segmentation effects according to hair styles or in case of containing noises and holes. In this study, the energy function of the test image is constructed according to the estimated prior distributions of hair location and hair color likelihood function. It is further optimized according to graph cuts algorithm and initial hair region is obtained. Finally, clustering algorithm and image post-processing techniques are applied to the initial hair region so that the final hair region can be segmented precisely. The proposed method is applied to MediaPipe hair segmentation pipeline.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.3
• /
• pp.11-20
• /
• 2024

Among the generative models in Artificial Intelligence (AI), especially Generative Adversarial Network (GAN) has been successful in various applications such as image processing, density estimation, and style transfer. While the GAN models including Conditional GAN (CGAN), CycleGAN, BigGAN, have been extended and improved, researchers face challenges in real-world applications in specific domains such as disaster simulation, healthcare, and urban planning due to data scarcity and unstable learning causing Image distortion. This paper proposes a new progressive learning methodology called Progressive Step Training (PST) based on the Auxiliary Classifier GAN (ACGAN) that discriminates class labels, leveraging the progressive learning approach of the Progressive Growing of GAN (PGGAN). The PST model achieves 70.82% faster stabilization, 51.3% lower standard deviation, stable convergence of loss values in the later high resolution stages, and a 94.6% faster loss reduction compared to conventional methods.

• Journal of the Korea Computer Graphics Society
• /
• v.20 no.2
• /
• pp.1-11
• /
• 2014

The quality of 2D-to-3D conversion depends on the accuracy of the assigned depth to scene objects. Manual depth painting for given objects is labor intensive as each frame is painted. Specifically, a human is one of the most challenging objects for a high-quality conversion, as a human body is an articulated figure and has many degrees of freedom (DOF). In addition, various styles of clothes, accessories, and hair create a very complex silhouette around the 2D human object. We propose an efficient method to estimate visually pleasing depths of a human at every frame in a monocular video. First, a 3D template model is matched to a person in a monocular video with a small number of specified user correspondences. Our pose estimation with sequential joint angular constraints reproduces a various range of human motions (i.e., spine bending) by allowing the utilization of a fully skinned 3D model with a large number of joints and DOFs. The initial depth of the 2D object in the video is assigned from the matched results, and then propagated toward areas where the depth is missing to produce a complete depth map. For the effective handling of the complex silhouettes and appearances, we introduce a partial depth propagation method based on color segmentation to ensure the detail of the results. We compared the result and depth maps painted by experienced artists. The comparison shows that our method produces viable depth maps of humans in monocular videos efficiently.

• Archives of design research
• /
• v.17 no.3
• /
• pp.111-122
• /
• 2004

The purpose of this study are two the first is to identify the positioning of product design and analyze its functional relationship with product development, manufacturing, marketing and sales in the comprehensive competitiveness evaluation model ; and the second is to estimate the determinants of QCI(quality competitiveness index), analyze the impact of product design on QCI, and compare the aforementioned results for the seven industrial sectors and the five product patterns. For this empirical analysis, this study surveyed 400 Korean manufacturing firms during August-October 2003. The major empirical findings are summarized as follows : First, the hypothesis on the positive effect of product design on QCI is accepted at a highly significant level (p < 0.001) for all : the manufacturing sector, seven industrial sectors, and five product categories. Second, the correlation analysis and factor analysis lead to the result that the effect of product design on QCI is estimated to be relatively very low, in comparison to those of product functionality and basic performance on QCI. These findings imply that Korean manufacturing sector has been still in the prematured stage at which product design has not played an important role yet. This study concludes that product design in line with other functions (product development, manufacturing, marketing, and sales) should make a good contribution to the improvement of QCI in the future.

• MISULJARYO - National Museum of Korea Art Journal
• /
• v.101
• /
• pp.48-65
• /
• 2022

Three Japanese Buddhist scuptures of Daehyusa temple in Gimcheon, Gyeongsangbuk-do province(慶尙北道) are enshrined in the main hall(本堂) of Henjoin temple in Daegu, an annex temple built by the Chisan School(智山派) of Shingi Shingon sect(新義眞言宗) in Daegu around 1910 when the Japanese colonial period began. The following statues are believed to have been made around this time: the Statue of Kobo daishi(弘法大師像), which is the figure of monk Kukai(空海, 774-835), the founder of the Shingon sect(眞言宗) of Japan, and the Statue of Dainichi nyorai(大日如來像) and the Statue of Fudo myo-o(不動明王像), which are notable sculptures representing the thought and belief of the Shingon sect. Most statues identified of Dainichi nyorai, Kobo daishi, and Fudo myo-o from the Japanese colonial period thus far are rock-carved statues or stone statues. The statues of Daehyusa temple are noteworthy in that they are the first discovered Japanese Buddhist sculptures made of wood, known to have been enshrined in Japanese temples. Furthermore, they are valuable sources that can provide clues to the religious atmosphere of the temples of Chisan School of Shingi Shingon sect at the time. Although these statues have formative features that partially reflect modern aesthetics, their iconographic origins date back to the Heian period(平安時代, 794-1185). In other words, the Statue of Dainichi nyorai inherits the religious tradition of the Statue of Dainichi nyorai, which was created by monk Kakuban(覺鑁, 1095-1143), the founder of the Shingi Shingon sect. The Statue of Kobo daishi follows the Shinnyo Shinno(眞如親王) style founded by Monk-Imperial Prince Shinnyo who was a disciple of monk Kukai. The Statue of Fudo myo-o manifests the Genjo(玄祖) style among the statues of Fudo myo-o. Although not much is known about how the statues were enshrined in the Henjoin temple in Daegu during the Japanese colonial period, it is very likely that these statues were created as Shingi Shingon sect's statue of three-wheeled body(三輪身). That is, it is estimated that the Statue of Dainichi nyorai, Statue of Kobo daishi, and Statue of Fudo myo-o have the characteristics of the body of self-nature chakra(自性輪身), the body of great dharma chakra(正法輪身), and the body of order chakra(敎令輪身), respectively. The fact that the statue of three-wheeled body was established in the Shingon sect in Japan by monk Kakuban also lends credence to this possibility. It is thought that people who came to the Henjoin temple in Daegu worshiped the statue of three-wheeled body to understand the teachings of the Dainichi nyorai. Although it is the case of Sildalsa Temple (悉達寺, the successor of Henjoin temple in Daegu) in the first half of the 1950s, the main hall includes features that reinforce the idea that the tradition of Kobo daisi faith(弘法大師信仰) is passed on. To illustrate, the human skeletal remains in the main hall of Sildalsa Temple reflect Koyasan(高野山)'s tradition in laying ashes to rest, which has been popular in conjunction with the Kobo daishi faith in Japan since the 11th century.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.32 no.12
• /
• pp.1939-1949
• /
• 2008

The purpose of this study was to identify differences or similarities across the ensembles of 2006 F/W fashion trends in the big fashion centers such as Paris, London, New York, and Seoul, by street fashion research. The study focuses on understanding of localized fashion trend in the marketplace. We used photograph observation and analyzed data by SPSS program. We found there is a significant difference in winter outfits at these different global fashion mega cities. Most Korean women were wearing light colored outer jackets and blue jeans were dominant style for pants. The majority of Paris/London, New York and Seoul people on the street were wearing wool/wool like coat. Padded coats were worn more by New Yorkers than by people in Seoul. For the bottom, there is a similarity between Paris/London, and New York City, in that skinny pants were popular. Koreans were wearing skinny pants mostly, but the percentage of mini skirts/shorts was also higher than any other cities. We found that the cross-cultural fashion mega trend is similar in clusters, but there is a slight difference of trend in clothing color, style and design details, and accessories by localized fashion cities. Not only direct observation but also identification of cultural characteristics and consumer behavior through the years will bring much more contributions to apparel industries.