• ETRI Journal
• /
• 제40권4호
• /
• pp.499-510
• /
• 2018

A vector space based augmented structural kinematic (VSASK) feature descriptor is proposed for human activity recognition. An action descriptor is built by integrating the structural and kinematic properties of the actor using vector space based augmented matrix representation. Using the local or global information separately may not provide sufficient action characteristics. The proposed action descriptor combines both the local (pose) and global (position and velocity) features using augmented matrix schema and thereby increases the robustness of the descriptor. A multiclass support vector machine (SVM) is used to learn each action descriptor for the corresponding activity classification and understanding. The performance of the proposed descriptor is experimentally analyzed using the Weizmann and KTH datasets. The average recognition rate for the Weizmann and KTH datasets is 100% and 99.89%, respectively. The computational time for the proposed descriptor learning is 0.003 seconds, which is an improvement of approximately 1.4% over the existing methods.

• Educational Technology International
• /
• 제9권1호
• /
• pp.1-23
• /
• 2008

Despite considerable commitment to the application of action learning as an organization development intervention, no identified systematic investigation of action learning practices has been reported. Based on a systematic literature review, the purpose of this paper is to identify whether researchers strike a balance between action and learning in their studies of action learning. Research findings in this study included: (1) only 32 empirical studies were found from the electronic database search; (2) based on the hypothesized continuum of Revans' original proposition of balancing action and learning, the author categorized 32 studies into three groups: action-oriented, learning-oriented, and balanced action learning; (3) there were only nine studies on balanced action learning among 32 empirical studies, whose insights included an effective use of project teams, applications of action learning for organization development, and key success factors such as time, reflection, and management support; (4) case study was among the most frequently used research method and only six quality studies met key methodological traits; and (5) therefore, more rigorous empirical research employing quantitative methods as well as case studies is needed to determine whether researchers strike a balance between action and learning in studies on action learning.

• The Korean Journal of Physiology and Pharmacology
• /
• 제11권3호
• /
• pp.85-88
• /
• 2007

Matrix metalloproteinases (MMPs) can degrade a wide range of extracellular matrix components. It has been reported that MMP-9 are activated after focal ischemia in experimental animals. (-)-Epigallocatechin-3-gallate (EGCG), a major constituent of green tea polyphenols, is a potent free radical scavenger and reduces the neuronal damage caused by oxygen free radicals. And it has been known that EGCG could reduce the infarction volume in focal brain ischemia and inhibit MMP-9 activity. To delineate the relationship between the anti-ischemic action and the MMP-9-inhibiting action of EGCG, we investigated the effect of EGCG on brain infarction and the activity of matrix metalloproteinase-9 induced by permanent middle cerebral artery occlusion (pMCAO) in ICR mice. EGCG (40 mg/kg, i.p. $15{\sim}30min$ prior to MCAO) significantly decreased infarction volume at 24 hr after MCAO. GM 6001 (50 mg/kg, i.p. $15{\sim}30min$ prior to MCAO), a MMP inhibitor, also significantly reduced infarction volume. In zymogram, MMP-9 activities began to increase at ipsilateral cortex at 2 hr after MCAO, and the increments of MMP-9 activities were attenuated by EGCG treatment. Western blot for MMP-9 also showed patterns similar to that of zymogram. These findings demonstrate that the anti-ischemic action of EGCG ire mouse focal cerebral ischemia involves its inhibitory effect on MMP-9.

• Preventive Nutrition and Food Science
• /
• 제16권4호
• /
• pp.291-298
• /
• 2011

Yam extracts (Dioscorea batatas) have been reported to possess a variety of functions. However, studies on its osteogenic properties are limited. In this study, we investigated the effect of ethanol and water extracts on osteoblast proliferation and bone matrix protein synthesis, type I collagen and alkaline phosphatase (ALP), using osteoblastic MC3T3-E1 cell model. MC3T3-E1 cells were cultured with yam ethanol and water extracts (0~30 mg/L) within 39 days of osteoblast differentiation period. Cell proliferation was measured by MTT assay. Bone matrix proteins were assessed by the accumulation of type I collagen and ALP activity by staining the cell layers for matrix staining. Also, the secreted (media) matrix protein concentration (type I collagen) and enzyme activity (ALP) were measured colorimetrically. Yam ethanol and water extracts stimulated cell proliferation within the range of 15~30 mg/L at 15 day treatment. The accumulation of type I collagen in the extracellular matrix, as well as secreted collagen in the media, increased with increasing doses of yam ethanol (3~15 mg/L) and water (3~30 mg/L) extracts. ALP activity was not affected by yam ethanol extracts. Our results demonstrated that yam extracts stimulated osteoblast proliferation and enhanced the accumulation of the collagenous bone matrix protein type I collagen in the extracellular matrix. These results suggest that yam extracts may be a potential activator for bone formation by increasing osteoblast proliferation and increasing bone matrix protein type I collagen. Before confirming the osteogenic action of yam, further studies for clarifying how and whereby yam extracts can stimulate this ostegenesis action are required.

• Computers and Concrete
• /
• 제12권1호
• /
• pp.19-36
• /
• 2013

In the finite element analysis of reinforced concrete structures, discrete representation of the steel reinforcing bars is considered advantageous over smeared representation because of the more realistic modelling of their bond-slip behaviour. However, there is up to now limited research on how to simulate the dowel action of discrete reinforcing bars, which is an important component of shear transfer in cracked concrete structures. Herein, a numerical model for the dowel action of discrete reinforcing bars is developed. It features derivation of the dowel stiffness based on the beam-on-elastic-foundation theory and direct assemblage of the dowel stiffness matrix into the stiffness matrices of adjoining concrete elements. The dowel action model is incorporated in a nonlinear finite element program based on secant stiffness formulation and application to deep beams tested by others demonstrates that the incorporation of dowel action can improve the accuracy of the finite element analysis.

• Kyungpook Mathematical Journal
• /
• 제51권3호
• /
• pp.261-281
• /
• 2011

In this paper, we will construct symmetric links by using the method adapted from the graph theory, and study a Seifert matrix of a symmetric link from the information of the Seifert matrix of the base link and the corresponding group action.

• 대한수학회보
• /
• 제54권5호
• /
• pp.1803-1825
• /
• 2017

Let X be a smooth scheme with an action of an algebraic group G. We establish an equivalence of two categories related to the corresponding moment map ${\mu}:T^{\ast}X{\rightarrow}g^{\ast}$ - the derived category of G-equivariant coherent sheaves on the derived fiber ${\mu}^{-1}(0)$ and the derived category of G-equivariant matrix factorizations on $T^{\ast}X{\times}g$ with potential given by ${\mu}$.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권2호
• /
• pp.832-854
• /
• 2019

Action recognition has been studied in computer vision field for years. We present an effective approach to recognize actions using a dimension reduction method, which is applied as a crucial step to reduce the dimensionality of feature descriptors after extracting features. We propose to use sparse matrix and randomized kd-tree to modify it and then propose modified Local Fisher Discriminant Analysis (mLFDA) method which greatly reduces the required memory and accelerate the standard Local Fisher Discriminant Analysis. For feature encoding, we propose a useful encoding method called mix encoding which combines Fisher vector encoding and locality-constrained linear coding to get the final video representations. In order to add more meaningful features to the process of action recognition, the convolutional neural network is utilized and combined with mix encoding to produce the deep network feature. Experimental results show that our algorithm is a competitive method on KTH dataset, HMDB51 dataset and UCF101 dataset when combining all these methods.

• 대한수학회보
• /
• 제50권2호
• /
• pp.667-674
• /
• 2013

Considering a special double-cover Q of the symmetric group of degree 3, we show that a proper non-regular approximate symmetry occurs from its quasigroup homogeneous space. The weak compatibility of any two elements of Q is completely characterized in any such quasigroup homogeneous space of degree 4.

• BMB Reports
• /
• 제42권9호
• /
• pp.561-567
• /
• 2009

Although many plant-derived phenolic compounds display antioxidant effects in biological systems, their mechanism of action remains controversial. In this study, the mechanism by which p-coumaric acid (p-CA) performs its antioxidant action was investigated in bovine aortic endothelial cells under oxidative stress due to high levels of glucose (HG) and arachidonic acid (AA), a free fatty acid. p-CA prevented lipid peroxidation and cell death due to HG+AA without affecting the production of reactive oxygen species. The antioxidant effect of p-CA was not decreased by buthionine-(S,R)-sulfoximine, an inhibitor of cellular GSH synthesis. In contrast, pretreatment with p-CA caused the induction of peroxidases that decomposed t-butyl hydroperoxide in a p-CA-dependent manner. Furthermore, the antioxidant effect of p-CA was significantly mitigated by methimazole, which was shown to inhibit the catalytic activity of 'p-CA peroxidases' in vitro. Therefore, it is suggested that the induction of these previously unidentified 'p-CA peroxidases' is responsible for the antioxidant effect of p-CA.