• 제목/요약/키워드: Short-cut method

검색결과 104건 처리시간 0.021초

The Difference of Perception of High School Girl Image according to Hair Style and Perceiver's Generation

  • Kim, Youngok;Lee, Eunsil
    • 패션비즈니스
    • /
    • 제18권6호
    • /
    • pp.208-221
    • /
    • 2014
  • The purpose of this study was to investigate the effects of the hair style, and perceiver's generation on the perception of a high schoolgirl's image. A quasi-experimental method by questionnaire was used. The subjects were 253 high schoolgirls and their mothers living in Jinju. The data were analyzed by using frequency, factor analysis, Cronbach's alpha-reliability coefficient, three-way ANOVA, and Duncan's multiple ranges test. The high schoolgirl's image was derived in five dimensions by factor analysis: schoolgirl look, prettiness, individuality, activeness, and cuteness. In the image based on the hair length, it was assessed that the short cut and the bobbed hair styles looked more like school girls. And the short cut and long hair styles were considered as showing individuality. On the other hand, the bobbed hair style was assessed as cuter than styles of other lengths. For the bangs, the short bangs were found to more school-girl look and cuter than the long bangs. The mother group, whose ages were older than that of the other group, evaluated that the stimulative pictures had more school-girl look, individuality, prettiness, activeness and cuteness, compared to how the school girl group evaluated. It was shown that the generation gap produced different opinions about the stimulative pictures. Therefore hair style, bangs style, and difference in generations were found to be the variables affecting perception of high school girl image.

적응적 오류 허용 라우팅 : SCP를 이용한 메쉬 구조에서의 RIFP 기법 개선 (Fault-Tolerant Adaptive Routing : Improved RIFP by using SCP in Mesh Multicomputers)

  • 정성우;김성천
    • 한국정보과학회논문지:시스템및이론
    • /
    • 제30권11호
    • /
    • pp.603-609
    • /
    • 2003
  • 다중 프로세서 환경에서의 오류 허용에 대한 적응적 라우팅 기법은 매우 중요한 요소이다. 특히 메쉬 구조를 갖는 다중 프로세서에서의 오류를 허용하는 라우팅 기법은 구현에 있어 간결함을 제공하는 환경으로, 다수의 오류를 허용하기 위해 직사각형 모양의 오류 블록으로 구성하여 라우팅을 수행한다. 이 경우, 블록 내부의 정상적인 노드 역시 오류 노드로 간주되어 전체 노드의 사용율을 저하시키는데, 오류 블록을 몇 개의 확장된 메쉬로 나누고, 구성된 확장된 메쉬들의 관계를 DAG(Directed Acyclic Graph)로 구성하고, 이 DAG에서의 확장된 메쉬간의 최단거리를 구하여 메시지를 전송하는 RIFP (Routing for Irregular Faulty Pattern)기법으로 이를 해결하였다. 그러나, 이 기법은 노드간에 주고받는 메시지가 거치는 hop의 수가 오히려 증가되는 문제가 발생하게 된다. 이러한 문제를 해결하기 위하여 본 논문에서는 증가되는 hop의 수를 억제하기 위해 목적 노드와 이웃 노드들로부터 오류 블록 경계 부분까지 직선 경로 SCP(Short-Cut Path)를 찾아 존재하는 경우, SCP를 적용하는 RIFP를 제안한다.

DeepLabCut과 Mask R-CNN 기반 반려동물 행동 분류 설계 (Design of Pet Behavior Classification Method Based On DeepLabCut and Mask R-CNN)

  • 권주영;신민찬;문남미
    • 한국정보처리학회:학술대회논문집
    • /
    • 한국정보처리학회 2021년도 추계학술발표대회
    • /
    • pp.927-929
    • /
    • 2021
  • 최근 펫팸족(Pet-Family)과 같이 반려동물을 가족처럼 생각하는 가구가 증가하면서 반려동물 시장이 크게 성장하고 있다. 이러한 이유로 본 논문에서는 반려동물의 객체 식별을 통한 객체 분할과 신체 좌표추정에 기반을 둔 반려동물의 행동 분류 방법을 제안한다. 이 방법은 CCTV를 통해 반려동물 영상 데이터를 수집한다. 수집된 영상 데이터는 반려동물의 인스턴스 분할을 위해 Mask R-CNN(Region Convolutional Neural Networks) 모델을 적용하고, DeepLabCut 모델을 통해 추정된 신체 좌푯값을 도출한다. 이 결과로 도출된 영상 데이터와 추정된 신체 좌표 값은 CNN(Convolutional Neural Networks)-LSTM(Long Short-Term Memory) 모델을 적용하여 행동을 분류한다. 본 모델을 바탕으로 행동을 분석 및 분류하여, 반려동물의 위험 상황과 돌발 행동에 대한 올바른 대처를 제공할 수 있는 기반을 제공할 것이라 기대한다.

단어 빈도와 α-cut에 의한 연관 웹문서 분류를 이용한 추천 시스템 (Recommendation System using Associative Web Document Classification by Word Frequency and α-Cut)

  • 정경용;하원식
    • 한국콘텐츠학회논문지
    • /
    • 제8권1호
    • /
    • pp.282-289
    • /
    • 2008
  • 협력적 필터링을 개선하기 위하여 많은 기술들이 개발되고 실용화되었으나 아이템의 연관 관계를 정확하게 반영하지는 못한다. 본 논문에서는 협력적 필터링의 문제점을 보완하기 위하여 단어 빈도와 ${\alpha}$-cut에 의한 연관 웹문서 분류를 이용한 추천 시스템을 제안한다. 제안된 방법은 형태소 분석을 통한 웹문서에서 단어를 추출하고 빈도 가중치를 계산한다. 추출된 단어를 Apriori 알고리즘을 이용해서 연관 규칙을 생성하고 신뢰도에 단어 빈도 가중치를 적용한다. 그리고 연관 규칙 하이퍼그래프 분할을 이용하여 연관 단어간의 유사도를 계산한다. 마지막으로 유사 클래스를 기반으로 연관 웹문서를 ${\alpha}$-cut을 이용하여 분류하고 개선된 코사인 유사도를 이용하여 유사도를 계산한다. 실험 결과 제안한 방법이 기존의 방법들보다 우수함을 확인하였다.

회분식 증류공정의 환류비 최적에 관한 연구 (A Study on Optimal Reflux Ratio for Batch Distillation)

  • 이영상;이의수
    • 제어로봇시스템학회논문지
    • /
    • 제8권10호
    • /
    • pp.833-841
    • /
    • 2002
  • We develope a model that can manipulate the reflux ratio of a batch distillation process in real time for optimal operations. Firstly, reflux ratio decision model for batch distillation unit was developed using the simple short-cut method. Secondly, more detailed rigorous method was applied to improve the accuracy of the model. Based on these models, operational strategies for the optimal reflux ratio was proposed. The results are illustrated with suitable examples and compared with the results using commercial simulator.

An Ensemble Cascading Extremely Randomized Trees Framework for Short-Term Traffic Flow Prediction

  • Zhang, Fan;Bai, Jing;Li, Xiaoyu;Pei, Changxing;Havyarimana, Vincent
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • 제13권4호
    • /
    • pp.1975-1988
    • /
    • 2019
  • Short-term traffic flow prediction plays an important role in intelligent transportation systems (ITS) in areas such as transportation management, traffic control and guidance. For short-term traffic flow regression predictions, the main challenge stems from the non-stationary property of traffic flow data. In this paper, we design an ensemble cascading prediction framework based on extremely randomized trees (extra-trees) using a boosting technique called EET to predict the short-term traffic flow under non-stationary environments. Extra-trees is a tree-based ensemble method. It essentially consists of strongly randomizing both the attribute and cut-point choices while splitting a tree node. This mechanism reduces the variance of the model and is, therefore, more suitable for traffic flow regression prediction in non-stationary environments. Moreover, the extra-trees algorithm uses boosting ensemble technique averaging to improve the predictive accuracy and control overfitting. To the best of our knowledge, this is the first time that extra-trees have been used as fundamental building blocks in boosting committee machines. The proposed approach involves predicting 5 min in advance using real-time traffic flow data in the context of inherently considering temporal and spatial correlations. Experiments demonstrate that the proposed method achieves higher accuracy and lower variance and computational complexity when compared to the existing methods.

Cinnamic acid, Cinnamaldehyde의 부위별 정량을 통한 계지(桂枝)의 음편(飮片) 규격 연구 (A Study of Cutting Methods by Comparing the Contents of Cinnamic acid and Cinnamaldehyde in Different Parts of Cinnamomi Ramulus)

  • 김영식;이금산;김정훈;최고야;정승일;조수인;주영승;김홍준
    • 대한본초학회지
    • /
    • 제26권2호
    • /
    • pp.11-15
    • /
    • 2011
  • Objectives : In the present study, the contents of cinnamic acid and cinnamaldehyde in three different parts of Cinnamomi Ramulus (CR) (the whole body, the bark part, and the wood part) was evaluated using UPLC (ultra performance liquid chromatography) in order to investigate a suitable cutting method. Methods : Analysis was performed on SMART LC with UV detector. Reference compounds were separated on Inertsil ODS-4 column ($2.1mm{\times}50mm$, $3{\mu}m$, GL Science, Japan) using isolation elution with water and acetonitrile each containing acetic acid at a flow rate of $500{\mu}L/min$. Additionally, samples of CR were purchased from pharmacy of medicinal herb. Results : The correlation coefficients of the cinnamic acid and cinnamaldehyde levels showed good linearity ($r^2{\geq}0.9999$) over the linear ranges. Furthermore, the bark part exhibited higher concentration levels of reference compounds than the wood part in all samples. In addition the bark exfoliation rates in oblique and perpendicular-long cut samples of CR were lower than the perpendicular-short cut samples. Conclusions : These results suggested that the optimal cutting method would be able to reduce the bark exfoliation. Therefore, the oblique or perpendicular-long cutting method is considered to be a better cutting type than the perpendicular-short cutting method.

Detection of short-term changes using MODIS daily dynamic cloud-free composite algorithm

  • Kim, Sun-Hwa;Eun, Jeong;Kang, Sung-Jin;Lee, Kyu-Sung
    • 대한원격탐사학회지
    • /
    • 제27권3호
    • /
    • pp.259-276
    • /
    • 2011
  • Short-term land cover changes, such as forest fire scar and crop harvesting, can be detected by high temporal resolution satellite imagery like MODIS and AVHRR. Because these optical satellite images are often obscured by clouds, the static cloud-free composite methods (maximum NDVI, minblue, minVZA, etc.) has been used based on non-overlapping composite period (8-day, 16-day, or a month). Due to relatively long time lag between successive images, these methods are not suitable for observing short-term land cover changes in near-real time. In this study, we suggested a new dynamic cloud-free composite algorithm that uses cut-and-patch method of cloud-masked daily MODIS data using MOD35 products. Because this dynamic composite algorithm generates daily cloud-free MODIS images with the most recent information, it can be used to monitor short-term land cover changes in near-real time. The dynamic composite algorithm also provides information on the date of each pixel used in compositing, thereby makes accurately identify the date of short-term event.

유전체가 입혀진 전자기 결합 동축 다이폴 어레이 안테나의 해석 (Analysis of Dielectric Coated Electromagnetically Coupled Coaxial Dipole Array (ECCDA) Antenna)

  • 구성모;양우석;이창원
    • 융합신호처리학회논문지
    • /
    • 제5권4호
    • /
    • pp.317-324
    • /
    • 2004
  • 전자기 결합된 동축 다이폴 안테나를 이론적으로 해석하였다. 방사도파관 모드와 퓨리어 변환을 이용하여 적분 방정식을 유도하였다. 적분 방정식에 포함된 적분들을 빠르게 계산하기 위하여 실수축을 따른 적분을 브랜치 컷 (branch cut)을 따른 적분 경로로 변환하여 계산하였다. 방사 특성에 미치는 슬롯과 다이폴 그리고 단락면의 길이 그리고 유전체 코팅의 영향을 조사하였다. 방시특성 또한 연구하였다. 제안된 방법의 타당성을 증명하기 위하여 기존에 발표된 방법으로 계산한 안테나의 방향성 이득과 비교하였고 또한 전력 보존 법칙을 만족시키는지를 확인하였다. 본 논문에서 제시한 방법으로 계산된 결과를 전자파 상용 시뮬레이터의 결과와 비교하여 잘 일치함을 확인하였다.

  • PDF

암석의 파괴인성 측정에서 나타나는 CB 시험편의 치수효과에 관하여 (The Size Effect in Measuring the Fracture Toughness of Rock using Chevron Bend Specimen)

  • 김재동;백승규
    • 터널과지하공간
    • /
    • 제2권2호
    • /
    • pp.251-264
    • /
    • 1992
  • In this study, the size effect in measuring the fracture toughness of rock was investigated using the ISRM Suggested Method for Fracture toughness using Chevron Bend Specimens. Total 58 specimens were prepared with 4 different diameters, 29, 42, 54, 68mm and center cut-chevron notch. In addition to this, to evaluated the effect of anisotropy of Jecheon granite, which is the sample for this study, core drilling direction was adjusted perpendicular(short transverse) and parallel(arrester) to the rift plane in the sample and the measured fracture toughness for each direction were compared. Important results obtained from this study are as follows. Level ll test condition is more adequate than l, because of low data scattering and precision and corrected fracture toughness of Jechoen granite measured and 2.2MPa{{{{ SQRT { m} }}}} for arrester direction with minimum initial crack length 0.7cm. From the relationship between core diameter and initial crack length presented in the ISRM testing method, the specimen diameter should be bigger than 47mm. The fracture toughnesses measured for arrester and short transverse directon show 10% difference. This is to the anisotropy of Jecheon granite possessing rift plane.

  • PDF