• 한국정보과학회논문지:소프트웨어및응용
• /
• 제28권12호
• /
• pp.947-954
• /
• 2001

본 논문에서는 인식된 얼굴의 포즈 상태 열에 오토마타 기법을 적용하여 다양한 헤드 제스처 를 인식하는 방법을 제안한다. 얼굴 영역의 추출에는 Yl7외 I성분인 최적의 얼굴색 정보와 적응적인 차영상 정보를 이용하며. 눈 영역 추출에는 소벨 연산자와 투영 기법. 그리고 눈의 기하학적 위치 정보를 이용 한다 얼굴의 상태 인식에는 계층적인 특징분석 방법을 사용하며, 인식된 얼굴 상태 열에 오토마타 기법을 적용하여 13가지 제스처; 준비, 상측, 하측, 좌측, 우측, 전진, 후퇴, 좌 윙크, 우 윙크, 좌 더블 윙크 우 더블 윙크, 긍정, 부정제스처를 인식한다. 총 8명으로부터 1,488 프레임의 영상을 취득하여 실험한 결과, 99.3%의 얼굴 영역 추출률 95.3%, 의 눈 영역 추출률, 94.1% 의 얼굴 상태 인식률. 그리고 93.3%의 헤드제 스처 인식률을 얻었다

• Coupled systems mechanics
• /
• 제10권6호
• /
• pp.509-520
• /
• 2021

Fiber-reinforced concrete (FRC) is a composite material where small fibers made from steel or polypropylene or similar material are embedded into concrete matrix. In a material model each constituent should be adequately described, especially the interface between the matrix and fibers that is determined with the 'bond-slip' law. 'Bond-slip' law describes relation between the force in a fiber and its displacement. Bond-slip relation is usually obtained from tension laboratory experiments where a fiber is pulled out from a matrix (concrete) block. However, theoretically bond-slip relation could be determined from bending experiments since in bending the fibers in FRC get pulled-out from the concrete matrix. We have performed specially designed laboratory experiments of three-point beam bending with an intention of using experimental data for determination of material parameters. In addition, we have formulated simple layered model for description of the behavior of beams in the three-point bending test. It is not possible to use this 'forward' beam model for extraction of material parameters so an inverse model has been devised. This model is a basis for formulation of an inverse model that could be used for parameter extraction from laboratory tests. The key assumption in the developed inverse solution procedure is that some values in the formulation are known and comprised in the experimental data. The procedure includes measured data and its derivative, the formulation is nonlinear and solution is obtained from an iterative procedure. The proposed method is numerically validated in the example at the end of the paper and it is demonstrated that material parameters could be successfully recovered from measured data.

• 비파괴검사학회지
• /
• 제15권2호
• /
• pp.395-406
• /
• 1995

초음파탐상시험을 통해 용접 결함의 종류를 정확히 구분하는 것은 정량적 비파괴시험을 위한 기본적인 단계로서 매우 중요한 문제인데, 이 문제는 최근 활발한 연구가 진행중인 초음파 형상 인식 기법의 적용에 의해 해결할 수 있다. 여기에서는 특징 추출, 특징 선택 그리고 결함 분류 등 초음파 형상 인식 기법의 세부 기술과 함께, 특히 최근 효율적인 분류기로 관심을 모으고 있는 확률 신경 회로망의 적용에 대해 논의하였다. 그리고 강 용접부 내부에 존재하는 결함을 균열, 기공, 슬래그 혼입의 3 종류로 분류하는 문제에 확률 신경 회로망을 적용한 예를 통하여, 초음파 형상 인식 기법의 효용성을 검증하였다. 또한 민감한 특징을 효율적으로 선별하는데 널리 사용되는 전방 특징 선택법과 그 적용에 대해서도 논의하였다.

• 복식문화연구
• /
• 제20권6호
• /
• pp.869-882
• /
• 2012

The purpose of this study is to obtain the patent for extracting the color map and color chip from the color image source and to develop color image map for fashion design. For this study, fashion image maps were produced from 210 pictures with Adobe Photoshop CS2 program targeting 200 university students from 2004 to 2006. The procedures for extracting the color map and color chip included providing the color image, the filtering phase, the segmentation phase, the extraction phrase, and the arrangement phase. Based on the results of this study, patent application was made to KIPO(Korean Intellectual Property Office) for this invention. The following effects can be expected from the standpoint of design based on the case study. First, it is a straight forward procedure to extract a color chip and color map from a color image. Second, it can be applied to various art works based on the recombination of colors as representative colors can be extracted from the related color image that combines a variety of colors. Third, desired colors can be selected based on the taste cluster classification or sensibility axis of design by extracting the representative color from the color image.

• ETRI Journal
• /
• 제42권1호
• /
• pp.90-100
• /
• 2020

Named entity recognition (NER) continues to be an important task in natural language processing because it is featured as a subtask and/or subproblem in information extraction and machine translation. In Urdu language processing, it is a very difficult task. This paper proposes various deep recurrent neural network (DRNN) learning models with word embedding. Experimental results demonstrate that they improve upon current state-of-the-art NER approaches for Urdu. The DRRN models evaluated include forward and bidirectional extensions of the long short-term memory and back propagation through time approaches. The proposed models consider both language-dependent features, such as part-of-speech tags, and language-independent features, such as the "context windows" of words. The effectiveness of the DRNN models with word embedding for NER in Urdu is demonstrated using three datasets. The results reveal that the proposed approach significantly outperforms previous conditional random field and artificial neural network approaches. The best f-measure values achieved on the three benchmark datasets using the proposed deep learning approaches are 81.1%, 79.94%, and 63.21%, respectively.

• 대한물리의학회지
• /
• 제12권3호
• /
• pp.49-58
• /
• 2017

PURPOSE: To summarize the evaluation tools of balance [Berg Balance Scale (BBS), timed up and Go (TUG), forward reaching test (FRT)], gait [6 m walking Test (6MWT)], and strength [Chair Stand Test (CST)] for patients with dementia. METHODS: The following databases were searched: Pub MED, Cochrane, Sciences Direct, and Web of Sciences. The inclusion criteria were as follows: 1) repeated measurement design, 2) subjects with dementia, 3) use of testing tools such as the BBS, TUG, FRT, 6MWT, and CST, 4) report the reliability. One reviewer performed the quality assessment of diagnostic accuracy study and two evaluators performed data extraction independently. RESULTS: Six articles and one letter were included. The interrater reliability of 6MWT, TUG, and CST, were acceptable (ICC>.90). However, FRT had unacceptable reliability. In test-retest reliability, only BBS has acceptable reliability (ICC>.90). Others had various reliabilities. The risk of interrater reliability bias was low in all studies. However, the risk of bias of intrarater reliability was low in five studies and moderate in two studies. CONCLUSION: The interrater reliability of the 6MWT, TUG, and CST were acceptable. However, in test-retest reliability, only BBS has acceptable reliability. Therefore, we suggest the use of BBS to test the balance of dementia patients. In addition, the study of tool reliability according to the subtype of dementia is needed in the future.

• Journal of Electrical Engineering and Technology
• /
• 제12권1호
• /
• pp.363-372
• /
• 2017

This paper presents a novel face recognition algorithm based on the deep convolution neural network and key point detection jointed local binary pattern methodology to enhance the accuracy of face recognition. We firstly propose the modified face key feature point location detection method to enhance the traditional localization algorithm to better pre-process the original face images. We put forward the grey information and the color information with combination of a composite model of local information. Then, we optimize the multi-layer network structure deep learning algorithm using the Fisher criterion as reference to adjust the network structure more accurately. Furthermore, we modify the local binary pattern texture description operator and combine it with the neural network to overcome drawbacks that deep neural network could not learn to face image and the local characteristics. Simulation results demonstrate that the proposed algorithm obtains stronger robustness and feasibility compared with the other state-of-the-art algorithms. The proposed algorithm also provides the novel paradigm for the application of deep learning in the field of face recognition which sets the milestone for further research.

• 한국지능시스템학회논문지
• /
• 제11권6호
• /
• pp.500-504
• /
• 2001

에미커 크기가 2$\times10\mu m^2$인 InGaP/GaAs이종접합 바이폴라 트랜지스터의 T자 모양으로 연결된 등기회로 요소를추출하기 위하여, 경계구간 설정이 개선된 유전자 알고리즘을 채택하였다. 이 소신호 모델 파리미터를 유전자 알고리즘을 사용하여, 다양한 순방향 바이서스에 측정한 S-파리미터로부터 추출하였다. 추출된 값들은 물리적인의미와 일관성을 보여준다. 모델 S-파리미터는 측정 S-파라미터와 2GHz-26.6GHz의 주파수 범위에서 잘 일치한다.

• 한국인공지능학회지
• /
• 제5권2호
• /
• pp.43-53
• /
• 2017

In this paper, the pattern of extracting the same expression is proposed by using the Android intelligent device to identify the facial expression. The understanding and expression of expression are very important to human computer interaction, and the technology to identify human expressions is very popular. Instead of searching for the emoticons that users often use, you can identify facial expressions with acamera, which is a useful technique that can be used now. This thesis puts forward the technology of the third data is available on the website of the set, use the content to improve the infrastructure of the facial expression recognition accuracy, in order to improve the synthesis of neural network algorithm, making the facial expression recognition model, the user's facial expressions and similar e xpressions, reached 66%.It doesn't need to search for emoticons. If you use the camera to recognize the expression, itwill appear emoticons immediately. So this service is the emoticons used when people send messages to others, and it can feel a lot of convenience. In countless emoticons, there is no need to find emoticons, which is an increasing trend in deep learning. So we need to use more suitable algorithm for expression recognition, and then improve accuracy.

• 한국지능시스템학회논문지
• /
• 제16권4호
• /
• pp.466-471
• /
• 2006

패턴 인식 문제에서 중요한 전처리 과정 중 하나는 특정을 선택하거나 추출하는 부분이다. 특정을 추출하는 방법으로는 PCA가 보통 사용되고 특정을 선택하는 방법으로는 SFS 나 SBS 등의 방법들이 자주 사용되고 있다. 본 논문은 진화 연산 방법으로써 비선형 최적화 문제에서 유용하게 사용되어 지고 있는 유전자 알고리즘을 특정 선택에 적용하는 유전자 알고리즘 특정 선택 (Genetic Algorithm Feature Selection: GAFS)방법을 개발하여 다른 특징 선택 알고리즘과의 비교를 통해 본 알고리즘의 성능을 관찰한다.