• Journal of IKEEE
• /
• v.27 no.3
• /
• pp.251-257
• /
• 2023

In this paper, we propose a dual-structured self-attention method that improves the lack of regional features of the vision transformer's self-attention. Vision Transformers, which are more computationally efficient than convolutional neural networks in object classification, object segmentation, and video image recognition, lack the ability to extract regional features relatively. To solve this problem, many studies are conducted based on Windows or Shift Windows, but these methods weaken the advantages of self-attention-based transformers by increasing computational complexity using multiple levels of encoders. This paper proposes a dual-structure self-attention using self-attention and neighborhood network to improve locality inductive bias compared to the existing method. The neighborhood network for extracting local context information provides a much simpler computational complexity than the window structure. CIFAR-10 and CIFAR-100 were used to compare the performance of the proposed dual-structure self-attention transformer and the existing transformer, and the experiment showed improvements of 0.63% and 1.57% in Top-1 accuracy, respectively.

• 한국사회복지학회:학술대회논문집
• /
• 1999.04a
• /
• pp.294-313
• /
• 1999

Volunteering has been defined as a form of altruistic helping behavior directed at improving other's welfare. Volunteering is, however, also identified as a type of collective action for community welfare. In this regard, this study tests whether neighborhood member's network and collective identity are more important determinants to explain participation in volunteer work than altruistic or normative motivation. This study estimates a model in which volunteering is determined by empathy, normative beliefs, neighborhood networks(friendship, contacts, and integration), and identity as a neighborhood member. This study shows that empathy, normative beliefs, and collective identity as a neighborhood member have significant impacts on participation in volunteer work. In addition, this study reveals that neighborhood member's network has an indirect impact on volunteering through identity as a neighborhood member. These results suggest that neighborhood community member's ties and collective identity are important sources for community welfare and collective volunteer work.

• Korean Journal of Social Welfare
• /
• v.38
• /
• pp.234-254
• /
• 1999

Volunteering has been defined as a form of altruistic helping behavior directed at improving other's welfare. Volunteering is, however, also identified as a type of collective action for community welfare. In this regard, this study tests whether neighborhood member's network and collective identity are more important determinants to explain participation in volunteer work than altruistic or normative motivation. This study estimates a model in which volunteering is determined by empathy, normative beliefs, neighborhood networks(friendship, contacts, and integration), and identity as a neighborhood member. This study shows that empathy, normative beliefs, and collective identity as a neighborhood member have significant impacts on participation in volunteer work. In addition, this study reveals that neighborhood member's network has an indirect impact on volunteering through identity as a neighborhood member. These results suggest that neighbor-hood community member's ties and collective identity are important sources for community welfare and collective volunteer work.

• Journal of the Architectural Institute of Korea Planning & Design
• /
• v.35 no.8
• /
• pp.65-72
• /
• 2019

The purpose of this study was to explore the influence of the topography and the neighborhood facility on the aged people's social network formation. The paper samples the aged people living in Changchun city located in a flat ground and Tonghua city located in a mountain region. Four selections of dwelling districts from the two cities are based on amount of the facilities maintaining the wealth and living area same to do the questionnaires. One hundred aged people are selected in the each dwelling district to fill out the questionnaires, and 386 valid questionnaires are finally obtained. The data obtained from the questionnaires are used to do some statistic analysis by using SPSS24.0 program. The results are shown as following: In the case of same amount of neighborhood facility for dwelling district, the aged people living in a flat ground use the facility more frequently and communicate more often with their friends and neighbors than the aged people living in a mountain region do, the social network forms more easily as well. In the other case of same topography dwelling districts, the aged people living in an abundant facility dwelling district use the facility more frequently and communicate more with their neighbors and friends than the aged people living in a mountain region do, so the social network forms more easily as well. Thus it can be seen that topography and amount of facility are significant influence factors of the aged social network formation.

• Journal of Communications and Networks
• /
• v.16 no.4
• /
• pp.371-377
• /
• 2014

Localized topology control is attractive for obtaining reduced network graphs with desirable features such as sparser connectivity and reduced transmit powers. In this paper, we focus on studying how to prolong network lifetime in the context of localized topology control for wireless multi-hop networks. For this purpose, we propose an energy efficient localized topology control algorithm. In our algorithm, each node is required to maintain its one-hop neighborhood topology. In order to achieve long network lifetime, we introduce a new metric for characterizing the energy criticality status of each link in the network. Each node independently builds a local energy-efficient spanning tree for finding a reduced neighbor set while maximally avoiding using energy-critical links in its neighborhood for the local spanning tree construction. We present the detailed design description of our algorithm. The computational complexity of the proposed algorithm is deduced to be O(mlog n), where m and n represent the number of links and nodes in a node's one-hop neighborhood, respectively. Simulation results show that our algorithm significantly outperforms existing work in terms of network lifetime.

• Smart Structures and Systems
• /
• v.28 no.6
• /
• pp.779-789
• /
• 2021

Cracks in railway sleeper are an inevitable condition and has a significant influence on the safety of railway system. Although the technology of railway sleeper condition monitoring using machine learning (ML) models has been widely applied, the crack recognition accuracy is still in need of improvement. In this paper, a two-stage method using edge detection and convolutional neural network (CNN) is proposed to reduce the burden of computing for detecting cracks in railway sleepers with high accuracy. In the first stage, the edge detection is carried out by using the 3×3 neighborhood range algorithm to find out the possible crack areas, and a series of mathematical morphology operations are further used to eliminate the influence of noise targets to the edge detection results. In the second stage, a CNN model is employed to classify the results of edge detection. Through the analysis of abundant images of sleepers with cracks, it is proved that the cracks detected by the neighborhood range algorithm are superior to those detected by Sobel and Canny algorithms, which can be classified by proposed CNN model with high accuracy.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.4
• /
• pp.432-436
• /
• 2010

Finding communities from network data including social networks can be done by clustering the nodes of the network as densely interconnected groups, where keeping interconnection between groups sparse. To exploit a clustering algorithm for community detection task, we need a well-defined similarity measure between network nodes. In this paper, we propose a new similarity measure named "Common Neighborhood Sub-graph density" and combine the similarity with affinity propagation, which is a recently devised clustering algorithm.

• Journal of Korean Public Health Nursing
• /
• v.29 no.3
• /
• pp.488-502
• /
• 2015

Purpose: This study was conducted to determine the neighborhood environmental factors associated with physical activity among adults living in rural communities. Methods: A cross-sectional descriptive survey was conducted with a convenience sample of 201 adults living in three Ri in Y-city, Gyeonggi-do. Data were collected from face-to-face interview by trained interviewers and were analyzed using a zero-inflated negative binominal regression model. Results: Participants reported engaged in moderate or vigorous physical activity was 76.1%; 10.5% of participants reported that they met moderate physical activity recommendations and 14.5% of participants reported that they met vigorous physical activity recommendations. Zero-inflated negative binominal regression analysis showed association of increasing days of physical activity with social cohesion (${\beta}=.130$, p=.005), social network (${\beta}=-.096$, p=.003), and safety for crime (${\beta}=-.151$, p=.036), and no days of physical activity was associated with no attainment of education and marginally associated with increasing BMI. Conclusion: Neighborhood environmental factors including social cohesion, social network, and crime for safety were significantly associated with physical activity of rural adults. Community health nurses should expand an approach for individual behavior change to incorporate rural adults' specific neighborhood environmental factors into physical activity interventions.

• 재활복지
• /
• v.21 no.1
• /
• pp.71-95
• /
• 2017

The purpose of this research is to suggest implications for the elderly with disability by taking a look at the influence of disability of elderly on successful aging and a moderating effect of social support network. This study analyzed the $4^{th}$ panel data from KReIS and the subjects for this research are 3,906 of elderly aged over 65 years. For the statistical analysis, the IBM SPSS Statistics v23.0 program was utilized. Technical statistics was verified, multiple regression analysis was used in order to take a look at the effects of moderating variable. The results of this research are as the following: Firstly, the disability of elderly had an negative effect on successful age. Secondly, as a social support network, friend & neighborhood network and family network had an positive effect on successful age. Thirdly, friend & neighborhood network among social support network had a moderating effect on successful aging. In other words, the more disabled the elderly are, the lower the level of successive aging became, it turned out that the bigger the social support of friend & neighborhood network, the less the negative effect of disability on successive aging. On the basis of the result, it was proposed a case management for the elderly with disability, assurance of mobility right, and self-help meeting program for social support network.

• Communications for Statistical Applications and Methods
• /
• v.18 no.2
• /
• pp.213-217
• /
• 2011

For the social network of n nodes, one might be interested in finding k nodes to disseminate the information as quickly as possible or to identify key nodes of high "local centrality". I propose two algorithms for determining k "local centers" of the network and work on a real case.