• Journal of the Korea Society for Simulation
• /
• v.31 no.1
• /
• pp.11-18
• /
• 2022

Recently, research to classify human activity using imagery has been actively conducted for the purpose of crime prevention and facility safety in public places and facilities. In order to improve the performance of human activity classification, most studies have applied deep learning based-transfer learning. However, despite the increase in the number of backbone network models that are the basis of deep learning as well as the diversification of architectures, research on finding a backbone network model suitable for the purpose of operation is insufficient due to the atmosphere of using a certain model. Thus, this study applies the transfer learning into recently developed deep learning backborn network models to build an intelligent system that classifies human activity using imagery. For this, 12 types of active and high-contact human activities based on sports, not basic human behaviors, were determined and 7,200 images were collected. After 20 epochs of transfer learning were equally applied to five backbone network models, we quantitatively analyzed them to find the best backbone network model for human activity classification in terms of learning process and resultant performance. As a result, XceptionNet model demonstrated 0.99 and 0.91 in training and validation accuracy, 0.96 and 0.91 in Top 2 accuracy and average precision, 1,566 sec in train process time and 260.4MB in model memory size. It was confirmed that the performance of XceptionNet was higher than that of other models.

• Journal of Korea Multimedia Society
• /
• v.21 no.12
• /
• pp.1459-1466
• /
• 2018

Robots which detect human and recognize action are important factors for human interaction, and many researches have been conducted. Recently, deep learning technology has developed and learning based robot's technology is a major research area. These studies require a database to learn and evaluate for intelligent human perception. In this paper, we propose a multi-modal sensor-based image database condition considering the security task by analyzing the image database to detect the person in the outdoor environment and to recognize the behavior during the running of the robot.

• Journal of the Korean Society of Safety
• /
• v.22 no.3 s.81
• /
• pp.81-87
• /
• 2007

In this study digital human models of ship construction tasks using modeling & simulation were constructed and human models' activities through human activity analysis were evaluated. Human Factors experts analyzed the actual workers' tasks using the same technique used in human activity analysis at the same time. The main objective of this study is to check a possibility of applying digital human modeling technique to ship construction tasks that are mostly non-standardized(not uniformed) whereas most applications of digital human modeling technique have been applied to standardized tasks. We evaluated postures of both real workers and digital humans by RULA. It turned out that the final scores of RULA evaluation on real workers are the same as the RULA scores for digital humans. However, there were differences of RULA detail scores between real workers and digital humans in the several processes related with the wrist twist and deviations. Those differences are considered to be resulted from the error in the on-site measuring worker's body dimension which could be reduced by accurate tools to correct data for body dimension and digital real drawings for facilities. The results showed possibility of application of digital human modeling and ergonomic analysis on informal work operations as well as formal operations in the shipbuilding industry.

• Electronics and Telecommunications Trends
• /
• v.33 no.3
• /
• pp.89-99
• /
• 2018

Human activity recognition (HAR) is a technology that aims to offer an automatic recognition of what a person is doing with respect to their body motion and gestures. HAR is essential in many applications such as human-computer interaction, health care, rehabilitation engineering, video surveillance, and artificial intelligence. Smartphones are becoming the most popular platform for activity recognition owing to their convenience, portability, and ease of use. The noticeable change in smartphone-based activity recognition is the adoption of a deep learning algorithm leading to successful learning outcomes. In this article, we analyze the technology trend of activity recognition using smartphone sensors, challenging issues for future development, and a strategy change in terms of the generation of a activity recognition dataset.

• Journal of Agricultural Extension & Community Development
• /
• v.22 no.4
• /
• pp.395-404
• /
• 2015

The purpose of this study was to examine the structural relationships between attributes, satisfaction, and behavioral intention of rural tourism activity. For this study, an onsite survey was conducted to tourists of 35 rural tourism farms. A total of 408 samples were collected for the analysis. The SPSS 20.0 and AMOS 18.0 statistical package were used for frequency, reliability, exploratory factor, confirmatory factor, and path analysis. Through exploratory factor analysis, four dimensions of rural tourism activity attribute were identified from 20 variables: 'human resource', 'activity', 'rural culture', and 'facility and environment'. The results of structural equation modelling indicate that 'human resource', 'activity', 'rural culture' had positive effect on behavioral intention. Through the results, agritourism farm operators will be able to develop effective marketing strategy and management skill.

• Bulletin of the Korean Chemical Society
• /
• v.31 no.10
• /
• pp.2854-2860
• /
• 2010

Synthesis of a new series of pyrimidinyl-imidazo[2,1-b]thiazole derivatives is described. Their antiproliferative activity against A375 human melanoma cell line was tested and the effect of substituents on the pyrimidinyl ring side chain was investigated. The biological results indicated that most of the newly synthesized compounds showed moderate activity against A375, compared with Sorafenib. Among all of these derivatives, the cyclic sulfamide derivatives IIIa, IIIb, and IIIe showed the most potent antiproliferative activity against A375 human melanoma cell line. The IC50 values of compounds IIIa,b were in nanomolar scale. In addition, compound IIIe ($IC_{50}=1.9\;{\mu}M$) also demonstrated more potent antiproliferative activity compared with Sorafenib ($IC_{50}=5.6\;{\mu}M$).

• Proceedings of the Korean Institute of Interior Design Conference
• /
• 2005.10a
• /
• pp.143-146
• /
• 2005

Human activity pattern has been changed as the contemporary urban society changes. Diverse activities repeat regular patterns as time passes. Diagram is a simple drawing which aims to organize and unify various information. The elements of the social behaviour could be spatialized by means of diagram applications. By using diagrams, architects understand contemporary urban society and form new space conditions. Time-based activity patterns consists of activity pattern in a restricted space and in urban structure for space use. Activity patterns for different time zones are explained by two types of diagrams, space occupation and flexibility of space, By the characteristic of space system structred by these diagrams, activities and programs are rearranged and variety of space is allowed through flexibility. Also, programs are mixed to apply to simultaneous occurrence of ever-changing human activities.

• Biomolecules & Therapeutics
• /
• v.7 no.1
• /
• pp.29-34
• /
• 1999

The cellular growth inhibition of 20 natural products was screened using SRB (sulforhodamine B) assay against 4 human cancer cell lines(SNU-1, SNU-C$_{4}$, Hep3B, Kato III). Ethanol extracts of Salvia miltiorrhiza, Saussurea lappa and Chelidonium majus showed potent anticancer activity among them, and further, it was fractionated into methylene chloride, hexane and methanol. Methylene chloride and methanol fraction of Salvia radix showed significant inhibitory activity against 4 human cancer cell lines. The effect of Salvia miltiorrhiza on anticancer activity in vitro models was evaluated with methylene chloride fraction of Salvia miltiorrhiza. Life span of ICR mice implanted with sarcoma-180 was increased by 40-61% and BDF$^{1}$ mice implanted with L1210 was increased by 66-89% upon intraperitoneal administration with methylene chloride fraction of Salvia miltiorrhiza. Based on these result, we suggested that Salvia miltiorrhiza showed anticancer activity on the in vivo and in vitro models

• Korean Journal of Pharmacognosy
• /
• v.50 no.2
• /
• pp.86-95
• /
• 2019

In the present study, the effect of Z. jujuba Miller var. inermis Rehder extracts on the secretion of atopic dermatitis (AD)-related cytokines and hyaluronidase activity was investigated. We prepared four fractions, butylene glycol (JB), ethanol (JE), and water (JW), with Z. jujuba Miller var. inermis extracts. JW significantly reduced the secretion of interleukin-8 and JE reduced the secretion of tumor necrosis factor-alpha in human keratinocyte HaCaT cells. Also, hyaluronidase activity was measured by enzyme assay and the fractions inhibited the activity in a dose-dependent manner. In addition, the human dermal fibroblast, HDF-n cells were treated with the extracts and antioxidant activities were measured. The results showed that the extracts increased the free radical scavenging activity and the superoxide dismutase activity. Taken together, Z. jujuba Miller var. inermis extracts reduced the secretion of AD-related cytokines and inhibited the hyaluronidase. In addition, the extracts showed antioxidant activity.

• Natural Product Sciences
• /
• v.24 no.4
• /
• pp.225-228
• /
• 2018

By activity-guided fractionation, gliotoxin was isolated as an antibacterial metabolite of the fungus Penicillium decumbens which was derived from the jellyfish Nemopilema nomurai. Gliotoxin was further evaluated for antibacterial activity against several piscine and human MDR (multidrug resistance) pathogens. Gliotoxin showed significant antibacterial activity against Gram-positive piscine pathogens such as Streptococcus iniae FP5228, Streptococcus iniae FP3187, Streptococcus parauberis FP3287, Streptococcus parauberis SPOF3K, S. parauberis KSP28, and Lactococcus garvieae FP5245. Gliotoxin showed strong activity especially against S. parauberis SPOF3K and S. iniae FP5228, which are resistant to oxytetracycline. It is noteworthy that gliotoxin effectively suppressed streptococci which are the major pathogens for piscine infection and mortality in aquaculture industry. Gliotoxin also showed strong antibacterial activity against multidrug- resistant human pathogens (MDR) including Enterococcus faecium 5270 and MRSA (methicillin-resistant Staphylococcus aureus) 3089.