• Journal of the Korea Society of Computer and Information
• /
• v.27 no.1
• /
• pp.69-75
• /
• 2022

This paper proposes a method of identifying the moment when a student with developmental disabilities needs assistance intervention in performing barista vocational training using virtual reality-based realistic contents. To this end, 21 students enrolled in a vocational training center for persons with disabilities were selected as study subjects. These students were trained to recognize the barista tools in a virtual reality environment. During the training, if students experienced difficulties and were unable to proceed further, they were asked to raise their hands or verbally request assistance. Using the collected data, two hypotheses were established based on the distance between the hand of the student and each barista tool in the virtual reality space in order to derive a criterion for judging the moment when an intervention is required. As a result of verifying the hypotheses, this study found that the cumulative distance from the hand of a student, who successfully finished the training without requiring an intervention, to the target barista tool as well as adjacent tools was significantly shorter than the cumulative distance to other barista tools.

• The KIPS Transactions:PartA
• /
• v.17A no.2
• /
• pp.103-112
• /
• 2010

Cluster sensor network is a sensor network where input nodes crowd densely around some nuclei. Steiner minimum tree is a tree connecting all input nodes with introducing some additional nodes called Steiner points. This paper proposes a mechanism for efficient construction of a cluster sensor network connecting all sensor nodes and base stations using connections between nodes in each belonged cluster and between every cluster, and using repetitive constructions of approximate Steiner minimum trees. In experiments, while taking 1170.5% percentages more time to build cluster sensor network than the method of Euclidian minimum spanning tree, the proposed mechanism whose time complexity is O($N^2$) could spend only 20.3 percentages more time for building 0.1% added length network in comparison with the method of Euclidian minimum spanning tree. The mechanism could curtail the built trees' average length by maximum 3.7 percentages and by average 1.9 percentages, compared with the average length of trees built by Euclidian minimum spanning tree method.

• The KIPS Transactions:PartA
• /
• v.14A no.5
• /
• pp.317-326
• /
• 2007

We propose more enhanced heuristic for the GOSST(Grade of Services Steiner Minimum Tree) problem in this paper. GOSST problem is a variation of Steiner Tree problem and to find a network topology satisfying the G-Condition with minimum network construction cost. GOSST problem is known as one of NP-Hard or NP-Complete problems. In previous our research, we proposed a heuristic employing Direct Steiner Point Locating strategy with Distance Preferring MST building strategy. In this paper, we propose new Steiner point locating strategy, Zigzag Steiner point Locating strategy. Through the results of out experiments, we can assert this strategy is better than our previous works. The Distance Zigzag GOSST method which hires the Distance Preferring MST building strategy and Zigzag Steiner point Locating strategy defrays the least network construction cost and brings 31.5% cost saving by comparison to G-MST, the experimental control and 2.2% enhancement by comparison to the Distance Direct GOSST method, the best GOSST method in our previous research.

• Culinary science and hospitality research
• /
• v.17 no.5
• /
• pp.15-29
• /
• 2011

In this study, I examined the problems of the current barista curriculum and sought how to improve them by feedback from domestic specialized baristas and comparative analyses of the training programs that take place in barista educational institutions. In order to identify problems in curriculum and what to be improved, I gained the opinions of 15 experts who are currently teaching students in barista educational institutions using Delphi technique. To sum up the opinions of experts, working level education of coffee has to be mainly done and substantial educations for employment such as cafe operating technique and marketing are needed. In addition, experts suggested enhancing service courses considering the nature of working in contact with customers and the course for baking goods that takes up a lot of sales. As a result of examining the current curriculum in the institutions for baristas, it seems necessary to supplement these training courses that experts suggested because the educations are very poor.

• The KIPS Transactions:PartA
• /
• v.15A no.2
• /
• pp.75-86
• /
• 2008

Black/white comic rendering is one of the researches in the field of non-photorealistic rendering(NPR). Black/white comics have been produced manually as yet. But these previous systems require lots of time and manual work. So we propose the COmics Rendering system on VIdeo Stream (CORVIS) which transforms video streams into black/white comic cuts. Stylized font, one of comic representations, can be used to express onomatopoeic words and mimetic dialogue exaggeratively. But current comic generation systems do not provide enough effects of stylized font. This paper proposes a model for stylized fonts to express various effects. Effects of stylized fonts we proposed include geometric deformations. Thus we could represent stylized fonts on the still cut of movies and the background texture on a cuts of plain black/white comics. The final quality of our system produced is good enough to compare with manual black/white comics.

• The Journal of the Korea Contents Association
• /
• v.19 no.9
• /
• pp.535-549
• /
• 2019

The purpose of this study was to examine what kind of images were presented on movie stars in the newspaper. For the purpose, we classified the time period according to the movie industry and media trend, selected representative stars by period, and collected 798 related articles reported in newspapers. As a result of analyzing the reporting trend, domestic and foreign topics, news format, and gender difference in collected movie star articles, it was found that the image of movie stars reproduced in newspaper articles had mostly neutral images that do not represent specific gender. Since the 2000s, news coverage was changed to reproduce various images rather than being fixed to particular images, and the subject of report became more diversified through comparison of domestic and foreign topics. In addition, articles in the form of book review decreased and the interview-type articles increased in number, and in the case of male movie stars, the proportion of articles based on works was high in comparison to female movie stars. This study has significance in that it explored the changes in the process of reproducing star images diachronically from the initial stage of stars to the modern times. And it is hoped that this study will serve as basic data for the follow-up studies on the process of reproducing various images in the multi-media era.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2010.04a
• /
• pp.111-114
• /
• 2010

스타 그래프, 버블정렬 그래프, 팬케익 그래프는 노드 대칭성(node symmetric), 최대 고장 허용도(maximum fault tolerance), 계층적 분할 성질을 갖고, 하이퍼큐브보다 망비용이 개선된 상호 연결망이다. 본 연구에서는 상호연결망으로 널리 알려진 스타 그래프와 그의 변형된 그래프들 사이의 임베딩 방법을 제안하고, 임베딩 비용을 분석한다. 임베딩 결과는 버블정렬 그래프는 스타 그래프에 연장율(dilation) 3, 스타그래프는 팬케익 그래프에 연장율 4로 각각 임베딩 가능하다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.432-434
• /
• 2022

우리는 전이 학습을 이용하여 원하는 특정 패션 스타일 분류기를 학습하였다. 패션 스타일 검색 결과물을 온라인 쇼핑몰과 연결하는 웹 서비스를 사용자에게 제공한다. 패션 스타일 분류기는 구글에서 이미지 검색을 통해 수집된 데이터를 이용하여 ResNet34[1]에 전이 학습하였다. 학습된 분류 모델을 이용하여 사용자 이미지로부터 패션 스타일을 17가지 클래스로 분류하였고 F1 스코어는 평균 65.5%를 얻었다. 패션 스타일 분류 결과를 네이버 쇼핑몰과 연결하여 사용자가 원하는 패션 상품을 구매할 수 있는 서비스를 제공한다.

• The KIPS Transactions:PartB
• /
• v.11B no.5
• /
• pp.555-562
• /
• 2004

A genre or a style is another view of documents different from a subject or a topic. The style is also a criterion to classify the documents. There have been several studies on detecting a style of textual documents. However, only a few of them dealt with web documents. In this paper we suggest sets of features to detect styles of web documents. Web documents are different from textual documents in that Dey contain URL and HTML tags within the pages. We introduce the features specific to web documents, which are extracted from URL and HTML tags. Experimental results enable us to evaluate their characteristics and performances.

• Journal of Broadcast Engineering
• /
• v.27 no.1
• /
• pp.104-123
• /
• 2022

Style transfer based on neural network provides very high quality results by reflecting the high level structural characteristics of images, and thereby has recently attracted great attention. This paper deals with the problem of resolution limitation due to GPU memory in performing such neural style transfer. We can expect that the gradient operation for style transfer based on partial image, with the aid of the fixed size of receptive field, can produce the same result as the gradient operation using the entire image. Based on this idea, each component of the style transfer loss function is analyzed in this paper to obtain the necessary conditions for partitioning and padding, and to identify, among the information required for gradient calculation, the one that depends on the entire input. By structuring such information for using it as auxiliary constant input for partition-based gradient calculation, this paper develops a recursive algorithm for super high-resolution image style transfer. Since the proposed method performs style transfer by partitioning input image into the size that a GPU can handle, it can perform style transfer without the limit of the input image resolution accompanied by the GPU memory size. With the aid of such super high-resolution support, the proposed method can provide a unique style characteristics of detailed area which can only be appreciated in super high-resolution style transfer.