• 한국멀티미디어학회논문지
• /
• 제23권2호
• /
• 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.

• 한국의류학회지
• /
• 제34권7호
• /
• pp.1184-1196
• /
• 2010

Several applications of body scanning technology have been commercialized or are currently under development. The virtual fit from 3D scans is most advanced form of virtual try-on. This article is an analysis of the comparison of user preferences for domestic versus foreign 3D virtual try-on systems. For this study, domestic i-Fashion Mall (www.ifashionmall.co.kr) and a Canadian company, My Virtual Model (www.mvm.com) were selected as the most representative online retailers that offer a virtual try-on system. The respondents were comprised of 70 Korean female college students in the age group 20-29. A five point Likert scale was used to evaluate the degree of the preference of virtual avatar and try-on images. T-test, cross table, and a chi-square independence test were conducted for data analysis. The results are as follow. 1. The representation about current looks according to each virtual fit image indicates that MVM is more accurate than i-Fashion Mall. 2. About decision confidence, respondents have decision confidence in i-Fashion Mall in the case of the avatar image; however, respondents have confidence in MVM or the fit image. 3. There were no significant differences in among waist size groups in accuracy, trust of each avatar image, while there were significant differences among waist size groups in the accuracy and trust of each virtual fit image. 4. About ease of use, respondents answered that i-Fashion Mall is superior to MVM. 5. The respondents prioritized the ‘fitting report’ of i-Fashion Mall and ‘Weight loss’ of MVM over other functionalities.

• 한국의류산업학회지
• /
• 제11권3호
• /
• pp.437-444
• /
• 2009

According to the activation of the fashion electronic commerce, this research investigated the merits and demerits and improvement plan of the 3D virtual try-on technology using the qualitative research method. This research was performed by interview with 70 evaluation group. The evaluation group of 3D virtual try-on was organized and the fit evaluation process by 3D human body scan and the 3D virtual try-on of the i-Fashion technology center were experienced. This study was performed by interview after the actual and virtual try-on about the casual shirt. The convenience and accuracy of measurement, usability in online shopping, body evaluation, complement of sizing system, and body shape management were discovered as merits. The requirement of high accuracy in sizing and avatar, limits of fabric expression, practical limitation by cost, vexatious of measurement garment, differences between real and virtual fittings, personal information leakage risk, and etc were pointed out as demerits. The mass customization, customized garment connected with medical service, humanized avatar, improved fitting report, entertainment, coordination, wardrobe manager were proposed as improvement plan.

• The International Journal of Costume Culture
• /
• 제13권2호
• /
• pp.118-129
• /
• 2010

Digital technology introduced into the clothing and fashion industry is evolving to digital virtual fashions and consumer-centered mass-customized production systems. Today the application of such 3D virtual try-on systems is being expanded gradually in the clothing industry. This study purposed to make virtual avatars and virtual garments using OptiTex and V-stitcher virtual software and compared the appearance of the virtual garments put on the virtual avatars. For this, we created virtual avatars and virtual garments using body measurements obtained from jive subjects of top jive body shapes, respectively, using $[TC]^2$ body scanner. According to the results of comparing the outcomes of the two different virtual software systems, virtual avatar II of V-Stitcher tended to have a more round and lifted hip and the waist line at a higher position. In addition, the body curves and shapes of a virtual avatar affect the appearance of virtual garments. This study applied the same body measurements to virtual avatars and the same pattern to virtual garments, but when different kinds of virtual software were used, the virtual avatars and virtual garments showed different appearance and fit. This result may mean that when customers buy apparel products using different kinds of virtual try-on systems, their evaluation of appearance can vary depending on the virtual try-on system. Therefore, research needs to be made actively for the development and use of linkage programs that can reflect actual body measurements between virtual software systems and 3D body scanning systems.

• 한국의류산업학회지
• /
• 제16권6호
• /
• pp.1008-1016
• /
• 2014

This study aims to analyze the changes in the mechanical properties of woven fabrics(cotton, linen, wool, silk, and polyester) by bonding fusible interlinings with varying deniers(10D, 20D, and 30D) for a 3D virtual try-on system(one that a user to try garments through screens using Avatar) developed over the last decade. We experimented with four mechanical properties and thicknesses of twenty-three specimens of interlining bonded fabrics including face fabrics and interlinings by using the KES-FB-AUTO system. The results showed that the tensile property increased(LT and RT increased and WT decreased) as the denier of the interlining increased; however, the change was slight. In contrast, the bending and shear properties increased significantly as the denier of the interlining increased on both the warp and the weft. This showed evidence that the interlining gives the fabrics size stability. The compression property was slight changed as the tensile property varies depending on the fibers and the denier of interlining. As expected, the thickness increased by bonding the interlining as the denier of interlining increased. From these results, we conclude that 3D users need to reflect these changes of woven fabrics by bonding interlinings when they try screen fittings to accurately express the fabric reality of manufactured garment.

• 한국의류학회지
• /
• 제34권10호
• /
• pp.1632-1646
• /
• 2010

This article is an analysis of the fit preference of the knitted jacket according to body types for 30's and 40's adult women through a comparison of real and virtual fit. A five point likert scale was used to evaluate the degree of the preference of real and virtual fit. The data were analyzed with t-test and ANOVA using statistical program SPSS 17.0. The results are as follow. 1. The real and virtual fit preference of the knitted jacket according to body types indicates that respondents preferred pattern 2 for body type N than the rest of the patterns; however, respondents preferred pattern 3 for body type A and H. 2. There were no significant differences between real and virtual fit preference. 3. It is necessary that the knitted jacket follow a similar size tolerance like a woven jacket. The results show that the next generation of virtual try-on systems need the development of a minus clothing ease scale.

• 복식문화연구
• /
• 제22권2호
• /
• pp.209-224
• /
• 2014

This study aims to compare the ease of shirt patterns in the men's wear market that have been constantly developing, and to analyze the fitness and appearance through the 3D virtual try-on system. The study selected three industrial patterns and two educational materials of classic-fit and slim-fit shirts for a total of ten items. The experiments involved virtual wearing, a stress contact point perspective map, and appearance evaluations. First, the ease differed significantly in the chest, waist, and sleeve cap height according to the patterns. Second, based on the stress, contact point, and transparency, there was ease in the order of Nam, D-brand, J-brand, Park, and S-brand in the classic-fit shirts while the order of D-brand, Nam, J-brand, Park, and S-brand resulted in the slim-fit shirts. Third, in the appearance evaluation, higher points resulted in the order of J-brand, Park, S-brand, D-brand, and Nam in both classic-fit and slim-fit. The results showed that proper ease differs greatly by shirt-fit and target age. Also, the shirts with more ease tended to be evaluated less favorably in the appearance evaluation. Based on the results, the study suggests that the proper ease in chest circumference is 18cm for the classic-fit shirts and 6 cm for the slim-fit shirts.

• 한국정보시스템학회지:정보시스템연구
• /
• 제31권4호
• /
• pp.145-160
• /
• 2022

Purpose The purpose of this study is to develop a virtual try-on deep learning model that can efficiently learn front and back clothes images. It is expected that the application of virtual try-on clothing service in the fashion and textile industry field will be vitalization. Design/methodology/approach The data used in this study used 232,355 clothes and product images. The image data input to the model is divided into 5 categories: original clothing image and wearer image, clothing segmentation, wearer's body Densepose heatmap, wearer's clothing-agnosting. We advanced the HR-VITON model in the way of Mixed-Precison, Gradient Accumulation, and sharing model weights. Findings As a result of this study, we demonstrated that the weight-shared MP-GA HR-VITON model can efficiently learn front and back fashion images. As a result, this proposed model quantitatively improves the quality of the generated image compared to the existing technique, and natural fitting is possible in both front and back images. SSIM was 0.8385 and 0.9204 in CP-VTON and the proposed model, LPIPS 0.2133 and 0.0642, FID 74.5421 and 11.8463, and KID 0.064 and 0.006. Using the deep learning model of this study, it is possible to naturally fit one color clothes, but when there are complex pictures and logos as shown in <Figure 6>, an unnatural pattern occurred in the generated image. If it is advanced based on the transformer, this problem may also be improved.

• 한국의류산업학회지
• /
• 제17권2호
• /
• pp.265-277
• /
• 2015

The aim of this study was to successfully reflect human body changes on the transformation of the virtual body within 3D virtual fitting spaces. For this purpose, existing problems of shape transformation of the virtual body were analyzed and regression equations which provides useful basic data for transformation of the virtual body that can be applied usefully to the 3D virtual fitting system was developed. Necessary data for the analyses were body measurement and 3D scan data of women with average physical form between the ages of 40 through 49. The reason that we used human body changes of the female subjects in their forties was based on the recognition that fundamental female body changes start to occur from age of forty. Body shapes were largely divided into 3 groups according to obesity which was found to be the biggest factor of shape change. Seven factors were extracted based on factor analysis of 47 body measurement categories and regression equations were created to extract specific measurements for each BMI group based on these seven factors. The major contribution of this paper can be summarized as follows. First, the regression equations to extract specific measurements based on the 7 representative variables remediated existing problem of virtual bodies as it increased the number of body shape transformation areas. Second, the regression equations helped to overcome the problem of current failing to reflecting changes in body cross-section shape based on simple girth measurements based on analysis of cross-section distances.

• 한국의류산업학회지
• /
• 제13권4호
• /
• pp.624-634
• /
• 2011

The purpose of this study is to develop the pattern of safety clothes used at flat or mountainous areas and to identify the pattern of safety clothes by conducting experimental evaluation of virtual wear. Three subjects were selected, based on fire fighters' physical constitution. A prototype design for safety clothing was determined after in-depth interviewing of professionals and surveying of Forest service staff and related agency. Wearing test should be carried out in the order of pattern making, virtual and real wearing evaluation. For data analysis, technical statistical values should be obtained by using body measurements of subject, frequency analysis and T-test. The jacket is designed to have a front extension and the entire length of clothing enough for wearer to put on it over ordinary shirts or sweater. The collar of jacket is of round type. Cyber reality enables to identify the movement and activity of virtual fitting model and to find out errors or problems in safety clothing prior to on-the-spot wear test, thus raising the precision level of pattern. There was significant difference between real and virtual fit preference. The results show that the virtual try-on system need the development of a specific style.