• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.5
• /
• pp.2253-2272
• /
• 2018

We propose a deep learning method for multi-focus image fusion. Unlike most existing pixel-level fusion methods, either in spatial domain or in transform domain, our method directly learns an end-to-end fully convolutional two-stream network. The framework maps a pair of different focus images to a clean version, with a chain of convolutional layers, fusion layer and deconvolutional layers. Our deep fusion model has advantages of efficiency and robustness, yet demonstrates state-of-art fusion quality. We explore different parameter settings to achieve trade-offs between performance and speed. Moreover, the experiment results on our training dataset show that our network can achieve good performance with subjective visual perception and objective assessment metrics.

• Proceedings of the KIEE Conference
• /
• 2005.04a
• /
• pp.152-154
• /
• 2005

This paper is proposed a fuzzy control and neural network based on the vector controlled induction motor drive system. The hybrid combination of fuzzy control and neural network will produce a powerful representation flexibility and numerical processing capability Also, this paper is proposed estimation and control of speed of Induction motor using fuzzy and neural network. The back propagation neural network technique is used to provide a real time adaptive estimation of the motor speed. This paper is proposed the experimental results to verify the effectiveness of the new method.

• Korean Security Journal
• /
• no.38
• /
• pp.83-107
• /
• 2014

As a part of school safety network construction project, a system of school marshal in Seoul-si and a project of academy safety guard in other regions are on the way of improvement. But, the crimes in and around the school, occurring frequently, are not solved easily, as this study points out. The effective improvement way of school safety network construction project is as followings. First, to enforce assistant-class school safety teacher system in order to nurture and educate experts in school safety, as a systematic support for strengthening occupational professionalism. Second, to enforce school safety network construction projects for construct circumstances through legal fix of school safety CPTED to make students free from crimes. This study looks into the real status and problem of school safety networks, and suggests system of school safety teacher and a legal enactment of CPTED, as systematic and a planed school safety countermeasure, not temporary.

• Journal of the Korean Institute of Educational Facilities
• /
• v.12 no.3
• /
• pp.13-21
• /
• 2005

The idea of the school networks for joint venture and use of school facilities is forming and using an organic network of schools that will communally share a group of facilities that are uniquely appropriate for each schools. Focused on two prominent cities of Northern Choongnam province, Cheonan and Asan city, the objective of this study is to present and exhibit the plans for joint venture and use of school facilities, and to use them as an example for other provincial cities within the nation where such renovation may be possible. The school networks is proposed based on units of 1 km radius around each site, that is named as small school network, in order to ease the communal use of the sidewalks by the residents and the facilities by the nearby schools. In Cheonan city, six small school networks were proposed, which will supply three to four elementary schools for an average of 50,000 people. On the other hand, Asan city, which showed shortages of elementary schools, three small school networks including the middle and high schools in addition to the elementary schools were proposed, that will provide three to six schools for an average of 30,000 people.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09a
• /
• pp.423-426
• /
• 2003

In this paper, we use the modular neural network and recurrent neural network structure to implement the artificial brain information processing. We also select related adaptive learning methods to learn the entirely new input in the existed neural network. With this, a part of information process in brain is implemented as and autonomous and adaptive model by neural network and further more, the entire model for information process in brain can be introduced.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2017.04a
• /
• pp.928-931
• /
• 2017

네트워크 트래픽 데이터를 정제하여, Convolutional Neural Network Model 훈련에 적합한 데이터 세트로 변환하는데, 그 방법은 패킷 단위의 트래픽 데이터를 이미지 형태로 만드는 것이다. 완성된 데이터 세트를 훈련데이터로 하여 Convolutional Neural Network Model에 훈련하고, 훈련데이터의 이미지 크기를 변환해가며 훈련시킨 결과에 대해 비교 분석 및 평가를 진행한다.

• 제어로봇시스템학회:학술대회논문집
• /
• 1999.10a
• /
• pp.100-103
• /
• 1999

In this paper, a method for the efficient implementation of the PICNET network configurator for a distributed control system(DCS) is proposed. The network configurator is composed of the time parameter estimator and the period scheduler, the file generator. The main role of network configurator estimates time parameter, the pre-run time scheduling of the user input and make the period transmission table for operating the PICNET based distributed control system.

• Proceedings of the Korean Powder Metallurgy Institute Conference
• /
• 2006.09a
• /
• pp.462-463
• /
• 2006

Monodispersed and nano-sized Ni powders were synthesized from aqueous nickel sulfate hexahydrate $(NiSO_4{\cdot}6H_2O)$ inside nonionic polymer network by using wet chemical reduction process. The sucrose was used as a nonionic polymer network source. The effect of reaction conditions such as the amount of sucrose and a various reaction temperature, nickel sulfate hexahydrate molarity. The influence of a nonionic polymer network on the particle size of the prepared Ni powders was characterized by means of X-ray diffraction (XRD), scanning electron microscopy (SEM), and particle size analysis (PSA). The results showed that the obtained Ni powders were strong by dependent of the reaction conditions. In particular, the Ni powders prepared inside a nonionic polymer network had smooth spherical shape and narrow particle size distribution.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.3
• /
• pp.1450-1463
• /
• 2019

The problem towards crowd behavior recognition in a serious clustered scene is extremely challenged on account of variable scales with non-uniformity. This paper aims to propose a crowed behavior classification framework based on a transferring hybrid network blending 3D res-net with inception-v3. First, the 3D res-inception network is presented so as to learn the augmented visual feature of UCF 101. Then the target dataset is applied to fine-tune the network parameters in an attempt to classify the behavior of densely crowded scenes. Finally, a transferred entropy function is used to calculate the probability of multiple labels in accordance with these features. Experimental results show that the proposed method could greatly improve the accuracy of crowd behavior recognition and enhance the accuracy of multiple label classification.

• Journal of Chest Surgery
• /
• v.45 no.4
• /
• pp.254-256
• /
• 2012

Chiari's networks are present in 1.5% to 4% of the population. They are a congenital disease characterized by a remnant of the right valve of sinus venosus and rarely have clinical significance. Chiari's network, as the name implies, has network-like shape, but there are other forms of appearance. We have experienced a case of a 60-year-old woman who had a cystic mass on the right atrium. Surgical treatment was performed forthe mass removal and differential diagnosis of the mass. There was no evidence of other tumor, but Chiari's network. As cystic form of Chiari's network have not been reported before, it is the first report of cystic form of Chiari's network.