• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.10
• /
• pp.2643-2657
• /
• 2023

Scene graphs are structured representations that can clearly convey objects and the relationships between them, but are often heavily biased due to the highly skewed, long-tailed relational labeling in the dataset. Indeed, the visual world itself and its descriptions are biased. Therefore, Unbiased Scene Graph Generation (USGG) prefers to train models to eliminate long-tail effects as much as possible, rather than altering the dataset directly. To this end, we propose Geometric and Semantic Improvement (GSI) for USGG to mitigate this issue. First, to fully exploit the feature information in the images, geometric dimension and semantic dimension enhancement modules are designed. The geometric module is designed from the perspective that the position information between neighboring object pairs will affect each other, which can improve the recall rate of the overall relationship in the dataset. The semantic module further processes the embedded word vector, which can enhance the acquisition of semantic information. Then, to improve the recall rate of the tail data, the Class Balanced Seesaw Loss (CBSLoss) is designed for the tail data. The recall rate of the prediction is improved by penalizing the body or tail relations that are judged incorrectly in the dataset. The experimental findings demonstrate that the GSI method performs better than mainstream models in terms of the mean Recall@K (mR@K) metric in three tasks. The long-tailed imbalance in the Visual Genome 150 (VG150) dataset is addressed better using the GSI method than by most of the existing methods.

• The Korean Journal of Applied Statistics
• /
• v.33 no.2
• /
• pp.115-122
• /
• 2020

K-means clustering uses a spherical or elliptical metric to group data points; however, it does not work well for non-convex data such as the concentric circles. Spectral clustering, based on graph theory, is a generalized and robust technique to deal with non-standard type of data such as non-convex data. Results obtained by spectral clustering often outperform traditional clustering such as K-means. In this paper, we review spectral clustering and show important issues in spectral clustering such as determining the number of clusters K, estimation of scale parameter in the adjacency of two points, and the dimension reduction technique in clustering high-dimensional data.

• Journal of the Korean Society for Library and Information Science
• /
• v.22
• /
• pp.331-360
• /
• 1992

This research presents author co-citation analysis of the subject area in the humanities - Korean history. Three approaches to multivariate analyses were used to display the inter-author relationships in the similarity matrix. Data on co-citation of sixty seven authors for the period of 1980­1989 were extracted from the database constructed by author. The author's name, here refers to a body of writings by a person, is the unit of analysis. The data were subjected to non-metric multidimensional scaling program create two-dimensional map of authors. Authors with similarity are clustered using hierarchical agglomerative procedure and it is found that five clusters in Korean history represent primarily research specializations. Author map of Korean history reveals the first dimension corresponding to subject orientation of authors and the second dimension corresponds to research method or research style. In factor analysis, each factor reflects research specialty made up of authors, and factor locadings demonstrate the breadth or concentration of sixty seven authors' scholarly contributions on Korean history. It is demonstrated that the· specific methodology employed by this research, author co-citation analysis, is useful to represent the intellectual structure of Korean history.

• Korean Journal of Biological Psychiatry
• /
• v.4 no.1
• /
• pp.67-73
• /
• 1997

The changes of electroencephalogram(EEG) in patients with dementia are most commonly studied by analyzing power or magnitude in certain traditionally defined frequency bands. However because of the absence of an identified metric which quantifies the complex amount of information, there are many limitations in using such a linear method. According to chaos theory, irregular signals of EEG can also result from low dimensional deterministic chaos. Chaotic nonlinear dynamics in the EEG can be studied by calculating the correlation dimension. The authors have analyzed EEG epochs from three patients with dementia of Alzheimer type and three matched control subjects. The multichannel correlation dimension is calculated from EEG epochs consisting of 15 channels with 16,384 data points per channel. The results showed that patients with dementia of Alzheimer type had significantly lower correlation dimension than non-demented controls on 12 channels. Topographic analysis showed that the correlation dimensions were significantly lower in patients with Alzheimer's disease on frontal, temporal, central, and occipital head regions. These results show that brains of patients with dementia of Alzheimer type have a decreased complexity of electrophysiological behavior. We conclude that the nonlinear analysis such as calculating correlation dimension can be a promising tool for detecting relative changes in the complexity of brain dynamics.

• Korean Journal of Veterinary Service
• /
• v.47 no.1
• /
• pp.1-7
• /
• 2024

Radiographic left atrial dimension (RLAD) is a valuable metric for assessing left atrial enlargement in dogs. While there have been studies on the use of RLAD and the increase in C-reactive protein (CRP) levels based on heart disease stages, there has been no prior research on the correlation between RLAD and CRP. In this study, the objective was to investigate the relationship between the rise in RLAD as myxomatous mitral valve disease (MMVD) stages advance and the increase in CRP levels with MMVD stage progression. In this study, a total of 30 small-breed dogs were included as subjects. These dogs were diagnosed with MMVD at the American College of Veterinary Internal Medicine (ACVIM) stage B1 or B2, or stage C, based on a comprehensive assessment including physical examination, thoracic radiography, and echocardiography. Measurements of VHS and RLAD were compared to assess any significant differences. There were significant differences in RLAD between dogs with MMVD ACVIM stage B1 and those with stage C. The monocytes and CRP levels showed significant differences between ACVIM stage B1, B2 and ACVIM C. Additionally, a significant correlation was observed between the RLAD and VHS measurements. This underscores the notable association between MMVD stage advancement and elevated monocyte and CRP levels. The RLAD scores exhibited a significant difference among dogs with ACVIM stages B1, B2, and C, and significant variations were also observed in monocyte and CRP levels. These results suggest that monocyte and CRP levels may be a valuable diagnostic indicator for heart disease in dogs during the diagnostic evaluation.

• 제어로봇시스템학회:학술대회논문집
• /
• 1993.10b
• /
• pp.524-529
• /
• 1993

Component Cost Analysis considers any given system driven by a white noise process as an interconnection of different components, and assigns a metric called "component cost" to each component. These component costs measure the contribution of each component to a predefined quadratic cost function. One possible use of component costs is for model reduction by deleting those components that have the smallest component cost. The theory of Component Cost Analysis is extended to include finite-bandwidth colored noises. The results also apply when actuators have dynamics of their own. When the dynamics of this input are added to the plant, which is to be reduced by CCA, the algorithm for model reduction process will be called Weighted Component Cost Analysis (WCCA). Closed-form analytical expressions of component costs for continuous time case, are also derived for a mechanical system described by its modal data. This is very useful to compute the modal costs of very high order systems beyond Lyapunov solvable dimension. A numerical example for NASA's MINIMAST system is presented.presented.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.47 no.11
• /
• pp.43-48
• /
• 2010

The main features of a dielectric-loaded surface plasmon polariton waveguide are analyzed such as mode effective index and propagation length. These parameters are calculated using the finite element method for different metal-polymer pairs while varying the ridge width and thickness. As a performance metric, we employ the 2D figure of merit including two conflicting parameters i.e. mode effective index and propagation length. The results obtained here allow one to identify the parameter range for realizing the dielectric-loaded surface plasmon polariton waveguide and to choose dimension and material of the ridge for subwavelength confinement and moderate propagation loss at telecom wavelengths.

• Bulletin of the Korean Mathematical Society
• /
• v.53 no.3
• /
• pp.875-884
• /
• 2016

Let x : $^{Mn-1}{\rightarrow}{\mathbb{R}}^n$ ($n{\geq}4$) be an umbilical free hyper-surface with non-zero principal curvatures. Then x is associated with a Laguerre metric g, a Laguerre tensor L, a Laguerre form C, and a Laguerre second fundamental form B, which are invariants of x under Laguerre transformation group. We denote the Laguerre scalar curvature by R and the trace-free Laguerre tensor by ${\tilde{L}}:=L-{\frac{1}{n-1}}tr(L)g$. In this paper, we prove a local classification result under the assumption of parallel Laguerre form and an inequality of the type $${\parallel}{\tilde{L}}{\parallel}{\leq}cR$$ where $c={\frac{1}{(n-3){\sqrt{(n-2)(n-1)}}}$ is appropriate real constant, depending on the dimension.

• Communications of the Korean Mathematical Society
• /
• v.9 no.2
• /
• pp.355-359
• /
• 1994

It would be interesting to know if energy minimizing harmonic maps between manifolds have symmetric properties when the manifolds under consideration have some. In this paper, we consider among others radial symmetry. A radially symmetric manifold M of dimension m is the one with a point, called a pole, and an O(m) action as an isometric rotation with respect to the pole, or more precisely a radially symmetric manifold M has a coordinate on which the metric is of the form $ds_{M}$$^2$ = d$r^2$ + m(r)$^2$d$\theta^2$ for some function m(r) depending only on r. Of course m(0) = 0, m'(0) = 1, and when m(r) = r, (M, $ds_{ M}$/$^2$) is the Euclidean space $R^2$.(omitted)

• Journal of the Korean Mathematical Society
• /
• v.40 no.3
• /
• pp.487-501
• /
• 2003

The paper is devoted to the description of changes of the structure of the holomorphic automorphism group of a bounded domain in \mathbb{C}^n under small perturbation of this domain in the Hausdorff metric. We consider a number of examples when an arbitrary small perturbation can lead to a domain with a larger group, present theorems concerning upper semicontinuity property of some invariants of automorphism groups. We also prove that the dimension of an abelian subgroup of the automorphism group of a bounded domain in \mathbb{C}^n does not exceed n.