• Annual Conference on Human and Language Technology
• /
• 1996.10a
• /
• pp.93-97
• /
• 1996

본 논문에서는 2 계층 분류기를 이용하여 일반적인 문서(보고서, 책, 잡지, 워드프로세서에서 출력 된 양식) 내의 다중 크기 및 다중 활자체의 인식을 위한 효과적인 방법을 제안하고 구현하였다. 다중언어 문자를 효과적으로 인식하기 위한 2 계층 분류기를 제안하였는데 이는 폰트 독립적 분류기와 폰트 의존적 분류기로 구성되어 있다. 제안된 방법의 성능 평가를 위하여 사무실에서 많이 사용하는 59 종류의 폰트와 각 폰트 당 3가지 크기의 글꼴과, 스캐너에서 지원되는 3가지 농도의 총 489개의 서로 다른 부류를 갖는 3,593,172 자를 대상으로 학습시킨 뒤에 일반 문서를 가지고 펜티엄 PC 상에서 인식 실험을 수행하였다. 실험 결과, 2계층 분류기를 갖는 시스템에서 96-98%의 인식률과 초당40자 이상의 인식 속도를 보여줌으로써 일반적인 문서에서 다중 크기 및 다중 활자체의 문자 인식에 매우 실용적인 가치가 있음을 확인했다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.10
• /
• pp.1323-1329
• /
• 2021

For hand expression recognition, hand pose recognition based on the static shape of the hand and hand gesture recognition based on the dynamic hand movement are used together. In this paper, we propose a hierarchical hand pose model based on finger position and shape for hand expression recognition. For hand pose recognition, a finger model representing the finger state and a hand pose model using the finger state are hierarchically constructed, which is based on the open source MediaPipe. The finger model is also hierarchically constructed using the bending of one finger and the touch of two fingers. The proposed model can be used for various applications of transmitting information through hands, and its usefulness was verified by applying it to number recognition in sign language. The proposed model is expected to have various applications in the user interface of computers other than sign language recognition.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.2
• /
• pp.85-94
• /
• 2017

This study examined the factors influencing the health status according to class and socioeconomic class recognition by the employment type. To take advantage of the 18 original sources of the Korea Labor panel materials carried out in the South Korea Labor Institute, 5,158 adults over 20 years old were included in the final analysis. The research results revealed that the incomes of regular workers and non-regular workers between the hierarchy and socioeconomic hierarchy recognition showed a statistically significant difference between the cage; it was consistent between the hierarchy in only the "heavy" category. Regular workers of society, and regardless of non-regular workers, were analyzed to be relatively low compared to the actual income. Regression analysis showed that regular jobs had higher socioeconomic hierarchy recognition. Non-regular workers had a lower income bracket and lower socioeconomic hierarchy recognition. In particular, in the case of non-regular workers, the pension was not subscribed and they had a poorer state of health. Therefore, the pension insurance payment for non-regular workers needs to compensate for the lost income during non-employment periods. In addition, the government should improve public relations through education, management fields, and cooperation with labor.

• Korean Journal of Gerontological Social Welfare
• /
• v.74 no.4
• /
• pp.125-155
• /
• 2019

This study used latent mean analysis and multi-group analysis to explore the differences in meaning in life and influencing factors based on Alderfer"s ERG theory by subjective social status. For this purpose, a survey was conducted on 998 older adults aged 65 and over using the welfare center and day care center in South Korea. The major findings are as follows. First, latent mean difference in subjective health by subjective social status was very large. In addition, the difference in the path coefficient of the meaning in life of subjective health by subjective social status was identified. Second, there was no significant difference in the belongingness between the subjective upper class and the middle class, whereas the subjective lower class was found to be significantly lower. On the other hand, belongingness was found to have a significant effect on the meaning in life at all levels. Third, there was no significant difference in engagement meaningful activity in all classes, and it had a positive effect on the improvement of meaning in life. Based on these results, this study suggests that practical and policy intervention plan to enhance meaning in life for older adults.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.10
• /
• pp.1977-1985
• /
• 2007

The character recognition technique is one of the artificial intelligence and has been widely applied in the automated system robot HCI(Human Computer Interaction), etc. This paper introduces the character set and the representative character that can be used in the recognition of the mage ROI. The character codes in this ROI include the digit, symbol, English and Hereat etc. We proposed the efficient multi-classifier structure by combining the small-size classifiers hierarchically. Moreover, we generated each small-size classifiers by delta-bar-delta learning algorithm. We tested the performance with various kinds of images and achieved the accuracy of 99%. The proposed multi-classifier showed the efficiency and the reliability for the mixed character code set.

• Information and Communications Magazine
• /
• v.31 no.3
• /
• pp.46-53
• /
• 2014

사용자-객체 상호작용을 위해서는 영상 내 객체의 종류와 위치를 정확하게 파악하여 사용자가 객체에 관련된 행동을 취할 경우, 그에 맞는 상호작용을 수행해야 한다. 이러한 객체인식에 널리 사용되는 지역 불변 특징량 기반의 방법론은 복잡한 배경이나 균일 물체에 대하여 잘못된 매칭으로 인식률이 저하된다. 본고에서는 이를 해결하기 위해, 컬러와 깊이 근접도 기반 깊이 계층을 나누고, 복잡 배경으로부터 생기는 잘못된 특징점 대응을 최소화 하기 위해 각 깊이 계층과 인식 물체 영상간의 특징점 대응을 수행한다. 또한, 각 깊이 계층영역에서 색상 히스토그램 재투영으로 객체의 위치를 추정하고 추정 영역과 인식 물체 영상간의 생상 및 깊이 유사도를 판단한다. 최종적으로, 복잡 배경 효과를 최소화한 특징점 대응의 수, 색상 및 컬러 유사도를 고려하여 신뢰도를 측정하여 객체를 인식하게 되며, 이를 통해 복잡한 배경에서도 사용자와 객체간의 유연한 상호작용이 가능해진다.

• Journal of KIISE:Software and Applications
• /
• v.36 no.12
• /
• pp.1028-1033
• /
• 2009

This paper presents a real-time hand gesture recognition approach for controlling a computer. We define hand gestures as continuous hand postures and their movements for easy expression of various gestures and propose a Two-layered Bayesian Network (TBN) to recognize those gestures. The proposed method can compensate an incorrectly recognized hand posture and its location via the preceding and following information. In order to vertify the usefulness of the proposed method, we implemented a Virtual Mouse interface, the gesture-based interface of a physical mouse device. In experiments, the proposed method showed a recognition rate of 94.8% and 88.1% for a simple and cluttered background, respectively. This outperforms the previous HMM-based method, which had results of 92.4% and 83.3%, respectively, under the same conditions.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.1
• /
• pp.54-61
• /
• 2009

This paper presents a new electromyogram (EMG) pattern recognition method based on the Hierarchical Temporal Memory (HTM) algorithm which is originally devised for image pattern recognition. In the modified HTM algorithm, a simplified two-level structure with spatial pooler, temporal pooler, and supervised mapper is proposed for efficient learning and classification of the EMG signals. To enhance the recognition performance, the category information is utilized not only in the supervised mapper but also in the temporal pooler. The experimental results show that the ten kinds of hand motion are successfully recognized.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1994.06c
• /
• pp.166-171
• /
• 1994

기존의 음성인식기들은 일반적으로 간단하면서도 성능이 우수한 계층별 학습에 의해서 설계된다. 계층별 학습은 통계적 패턴인식에서의 ML 추정기법처럼 모델간의 독립성이 보장되고 무한한 양의 학습데이타가 주어진다는 가정에 기초하고 있다. 그러나, 대상어휘집합에 음운학적으로 유사한 어휘가 많이 포함되어 있는 인식문제에 있어서는 모델간의 독립성이 보장되지 못하고, 실제 주어지는 grktmqepdlk의 양도 제한되므로 기존의 합습알고리즘에는 한계가 있다. 따라서 본 논문에서는 그러한 가정상의 문제점으로 생기는 인식기의 성능저하를 개선할 수 있는 변별력 있는 학습알고리즘들을 검토하고 그의 일반적인 접근방법들에 대해서 논의한다.

• Journal of KIISE:Information Networking
• /
• v.37 no.1
• /
• pp.57-65
• /
• 2010

Recently, the character recognization technique is strongly needed to enable the mobile communication devices with cameras to gather input information from the users. In general, it is not easy to reuse a CBOCR(Camera Based Optical Character Recognizer) module because of its dependency on a specific platform. In this paper, we propose a software architecture for CBOCR module providing the easy adaptability to various mobile communication platforms. The proposed architecture is composed of the platform dependency support layer, the interface layer, the engine support layer, and the engine layer. The engine layer adopts a plug-in data structure to support various hardware endian policies. We show the effectiveness of the proposed method by applying the architecture to a practical product.