• International Journal of Internet, Broadcasting and Communication
• /
• v.9 no.4
• /
• pp.1-7
• /
• 2017

Clipart is artificial visual contents that are created using various tools such as Illustrator to highlight some information. Here, the style of the clipart plays a critical role in determining how it looks. However, previous studies on clipart are focused only on the object recognition [16], segmentation, and retrieval of clipart images using hand-craft image features. Recently, some clipart classification researches based on the style similarity using CNN have been proposed, however, they have used different CNN-models and experimented with different benchmark dataset so that it is very hard to compare their performances. This paper presents an experimental analysis of the clipart classification based on the style similarity with two well-known CNN-models (Inception Resnet V2 [13] and VGG-16 [14] and transfers learning with the same benchmark dataset (Microsoft Style Dataset 3.6K). From this experiment, we find out that the accuracy of Inception Resnet V2 is better than VGG for clipart style classification because of its deep nature and convolution map with various sizes in parallel. We also find out that the end-to-end training can improve the accuracy more than 20% in both CNN models.

• The KIPS Transactions:PartB
• /
• v.8B no.5
• /
• pp.565-572
• /
• 2001

Analyzing functions with small values of the product of position and frequency uncertainties have many advantages in image processing and data compression. Until now, this values has been computed based on the uncertainty principle, but the computed frequency uncertainty is not practical the human visual filters which have on-zero peak response frequencies. A new metric for the frequency uncertainty is used to calculate a deviation about the frequency which has maximum response. The joint effective widths for various functions are derived. As the result of analysis, the joint uncertainty for many functions converges to 0.5 as the joint parameter increases. Furthermore. Gabor cosine function shows an excellent performance among the mentioned functions.

• 한국HCI학회:학술대회논문집
• /
• 2008.02b
• /
• pp.392-398
• /
• 2008

이 연구는 정보의 효과적인 전달을 목적으로 하는 정보디자인에 있어서 지식의 활용에 대해 고찰하며, 정보시각화와 지식시각화의 특성에 대한 비교분석을 목적으로 한다. 지식은 일종의 고부가 가치의 정보로서 주어진 데이터에 관한 인간의 인지활동을 내포하고 있다. 지식은 정보에 관한 사용자의 해석, 인지, 이해 등을 거친 것으로서 정보 이상의 가치를 지닌 것으로 사용자의 심상과 인지를 확장시켜 새로운 정보와 지식의 생성 및 의사 결정에 사용될 수 있다. 지식의 표현은 현재 정보의 양적 팽창으로 인해 정보디자인에서 나타나는 정보과중, 해석오류와 오역, 사용자 태도 등의 문제를 해결하는 대안이 될 수 있다. 지식시각화는 지식의 전달과 장조를 위한 방법론으로써 사람들이 알고 있는 것을 좀 더 풍부한 의미로 전달한다. 또한 개인 또는 그룹 사이의 지식의 장작을 촉진하고 전달을 개선하며 습득을 용이하게 할 수 있다. 이 연구는 지식시각화란 정보디자인의 새로운 연구 분야로써 이론적 부분을 정립하고 표현방법 및 효과적인 프로세스 모델 설계를 연구하였다. 정보시각화와 지식시각화의 이론적 특성과 표현 방법 요소의 비교분석을 통해 지식시각화의 성격과 특정을 파악하였다. 또한 지식시각화의 대표적 표현방법을 도출하고, 지식의 표현에 적함한 시각화 방법 에 관해 고찰하였다. 더불어 지식의 실질적 표현에 기초가 되는 지식 구조화에 적합한 시각화 프로세스 모델을 제안하였다. 일반적인 시각화에서는 다양한 관점 제시나 지원구분을 하고 있지 않기 때문에 이 연구를 통해 도출되는 시각화 요소를 이용한 합리적인 시각화 프로세스 모델은 지식의 표현에 있어 효과적일 것이라 생각한다.

• Journal of Communications and Networks
• /
• v.12 no.2
• /
• pp.168-173
• /
• 2010

Error concealment techniques are significant due to the growing interest in imagery transmission over error-prone channels. This paper presents a spatial error concealment technique for losslessly compressed images using least significant bit (LSB)-based data hiding to reconstruct a close approximation after the loss of image blocks during image transmission. Before transmission, block description information (BDI) is generated by applying quantization following discrete wavelet transform. This is then embedded into the LSB plane of the original image itself at the encoder. At the decoder, this BDI is used to conceal blocks that may have been dropped during the transmission. Although the original image is modified slightly by the message embedding process, no perceptible artifacts are introduced and the visual quality is sufficient for analysis and diagnosis. In comparisons with previous methods at various loss rates, the proposed technique is shown to be promising due to its good performance in the case of a loss of isolated and continuous blocks.

• Proceedings of the KIEE Conference
• /
• 2004.07d
• /
• pp.2631-2633
• /
• 2004

This paper describes the construction of a half sphere screen driving tilting simulator that can perform six degree-of-freedom( DOF) motions simmulator to a tilting train. The mathematical equations of Tilting Train dynamics are first derived from the 6-DOF bicycle model and incorporated with the bogie. carbody, and suspension subsystems. The equations of motion are then programmed by visual C++ code. To achieve the simulator functions. a motion platform that is constructed by six electric-driven actuators is designed. and its kinetics/inverse kinetics analysis is also conducted. Driver operation signals such as carbady angle, accelerator. and tilting positions are measured to trigger the Tilting dynamics calculation and further actuate the cylinders by the motion platform control program. In addition. a digital PID controller is added to achieve the stable and accurate displacements of the motion platform. The experiments prove that the designed simulator is adequate in performing some special rail mad driving situations discussed in this paper.

• International Journal of Human Ecology
• /
• v.7 no.2
• /
• pp.29-41
• /
• 2006

Various current events provide evidence that society is undergoing changes in perceptions of social relationships. Specifically, visual media in the form of advertisements can convey images which reflect society's values and concepts about role relationships. The purpose of this research was to examine ads in fashion magazines to determine what types of model roles and role relationships typically appear in fashion advertising which can mirror society's values. A content analysis was conducted of ads obtained from US Vogue and US GQ for the year 2002. Six kinds of roles/relationships were found: (1) Narcissism (representing self absorption), (2) sexually enticing opposite-sex relationships, (3) close/romantic same-sex relationships, (4) friend relationships, (5) family relationships, and (6) independent relationships. Of these, narcissism predominated, however, a small number of sexually provocative ads appeared as well as same-sex romantic relationships. Because sole (single) models were more typical, they also were examined to determine ways in which they relate to the audience. Characteristics examined included body presentation & pose, eye gaze, and facial expression. Direct eye gaze was the typical way to engage the audience. Gender differences were apparent: smiling was more typical of women, indifference for men. The symbolic meaning and values investigated from this research are the blurring of gender identity portrayed in homosexual imagery, family values, and the value of youth. The consistency of models' race in ads does not portray the diversity reflected in the demographic census.

• The Bulletin of The Korean Astronomical Society
• /
• v.42 no.2
• /
• pp.43.4-44
• /
• 2017

Bar fractions depend on the properties of host galaxies. However, the observational studies did not provide consistent tendency. We investigated the bar fractions and their dependence on properties of host galaxies using three bar classifications: visual inspection, ellipse fitting method and Fourier analysis from a volume-limited sample of 1,698 disk galaxies brighter than Mr=-15.2 within z = 0.01 from the Sloan Digital Sky Survey (SDSS) Data Release 7 (DR7). We found two causes to make the discrepancy in previous studies. One is caused by the difficulty in automatically identifying bars for bulge-dominated galaxies. In particular, ellipse fitting methods could miss early-type barred galaxies when a large bulge weakens the transition between a bar and disk. The other is caused by the difference in the correlation between the bar types and host morphology for strong bars and weak bars. Strong bars are preponderant in early-type spirals which are red, bulge-dominated and highly concentrated, whereas weak bars are frequent in late-type spirals which are blue, disk-dominate and less-concentrated. Therefore, how much weak bars they contain affects the trend of bar fraction on host galaxy properties. We also discuss the effect of host properties on the formation, evolution, and destruction of bars.

• Physical Therapy Korea
• /
• v.14 no.4
• /
• pp.58-65
• /
• 2007

The purpose of this study was to investigate the effects of serratus anterior strengthening exercises on scapular position and Visual Analog Scale (VAS) pain measurements taken at the resting position in young adults with adducted scapular. The exercise program included stretching of the scapular retractor and strengthening of the serratus anterior muscle. We measured the distance from the midline of the thorax to the vertebral border of the scapular with a tape line (Superior Kibler), and the distance from the 7th cervical spinous process to the acromial angle with 3-dimension motion analysis system, to compare the resting scapular position before and after exercise. Fifteen subjects with adducted scapular were recruited to compare the resting scapular position and VAS. The distance from 7th cervical spine process to acromial angle of the scapular and VAS decreased significantly (p<.01) after exercise, while the distance from the midline of the thorax to vertebral border of the scapular increased (p<.05). The conclusion is that the serratus anterior exercise program altered the resting scapular position and decreased VAS.

• Proceedings of the Korean Society for Bioinformatics Conference
• /
• 2004.11a
• /
• pp.296-306
• /
• 2004

LC/MS를 이용하여 펩타이드 혹은 단백질 같은 물질을 분석하는 실험이 급격히 늘어남에 따라 LC/MS 데이터를 자동으로 처리하는 기술에 대한 요구가 커지고 있다. 이러한 LC/MS 데이터의 자동 분석 기술에 대한 연구는 현재 활발히 진행되어 왔고, 이를 직접 구현한 여러 상용 소프트웨어들이 개발되어 있는 상태이다. LC/MS 데이터는 noise 제거, background 데이터 제거, deconvolution 알고리즘을 적용한 분자량(molecular weight) 할당 등의 작업을 거쳐 분석하게 된다. 이러한 과정을 거쳐 얻어진 분자량에 대한 데이터가 올바른 값인지 검증하는 작업이 필요하다. 본 논문에서는 이러한 검증 작업과 관련하여 Peak Clustering and Fitting(이하 PC&F)에 대한 알고리즘을 제안한다. PC&F은 peak 데이터들이 지니고 있는 속성에 대한 Mahalanobis distance를 이용하여 peak 데이터를 각 retention time에 따라 clustering 분석을 하는 작업이다. 본 논문에서 제안하는 PC&F 알고리즘을 Microsoft Visual C++ 6.0 MFC 환경에서 직접 개발한 소프트웨어(PeakClusterFitLCMS)로 실험하였다. 실험결과 PC&F 작업을 통해 동일한 구성물질로부터 발생한 peak 데이터를 모아서 보다 신뢰할 수 있는 분자량을 구할 수 있었고, 구성물질에 의해 발생되지 않은 noise peak 데이터를 찾아 제거시킬 수 있음을 확인할 수 있었다.

• 한국전산유체공학회:학술대회논문집
• /
• 2010.05a
• /
• pp.347-351
• /
• 2010

In this paper, our effort to apply 3-D Virtual Reality system for stereoscopic visualization of flow data is briefly described. This study is an extension of our previous and on-going research efforts to develop DATA(Data Analysis and Visualization Application) program, which is a data visualization program developed by using Qt as GUI development environment and OpenGL as graphic library. The program is developed upon the framework of object-oriented programming and it was originally developed by using Qt 3.3.3 environment. In this research the program is converted into a Qt 4.3.3-compatible version, and this new version is developed on Visual Studio 2005. And to achieve a stereoscopic viewing capability, two graphic windows are used to render its own viewing image for the lift and right eye respectively. These two windows are merged into one image using 3D monitor and the viewers can see the data visualization results with stereoscopic depth effects by using polarizing glasses. In this paper three dimensional data visualization with stereoscopic technique combined with 3D Monitor is demonstrated, and the current achievement would be a good start-up for further development of low-cost high-quality stereoscopic data visualization system.