• Journal of KIISE
• /
• v.43 no.12
• /
• pp.1356-1364
• /
• 2016

This paper proposes a novel aesthetic photo recomposition method using a deep convolutional neural network (DCNN). Previous recomposition approaches define the aesthetic score of photo composition based on the distribution of salient objects, and enhance the photo composition by maximizing the score. These methods suffer from heavy computational overheads, and often fail to enhance the composition because their optimization depends on the performance of existing salient object detection algorithms. Unlike previous approaches, we address the photo recomposition problem by utilizing DCNN, which shows remarkable performance in object detection and recognition. DCNN is used to iteratively predict cropping directions for a given photo, thus generating an aesthetically enhanced photo in terms of composition. Experimental results and user study show that the proposed framework can automatically crop the photo to follow specific composition guidelines, such as the rule of thirds.

• IEIE Transactions on Smart Processing and Computing
• /
• v.5 no.1
• /
• pp.1-4
• /
• 2016

In this paper, we present an algorithm for adjusting the position of a user-specified object considering image aesthetics. Specifically, the user-specified object is positioned according to the rule of thirds by inserting or deleting unimportant seam lines from the image. To find such seam lines, a novel weight map is designed using the spatial and color distances from the object. We also design and analyze two approaches to seam carving and insertion. Experimental results show that the proposed method can be used as an effective semi-automatic image recomposition scheme.

• Journal of the Korean Society of Industry Convergence
• /
• v.3 no.4
• /
• pp.369-378
• /
• 2000

Conventional fractal image compression methods have many problems in searching time for matching domain block. Proposed method is an improved method of Fisher's Quadtree Decomposition in terms of time, compression ratio, and PSNR. This method determines range block in advance using QR algorithm. First, input image is partitioned to $4{\times}4$ range block and then recomposition is performed from bottom level to specified level. As a result, this proposed method achieves high encoding and decoding speed, high compression ratio, and high PSNR than Fisher's Quadtree Decomposition method.

• Research in Mathematical Education
• /
• v.6 no.1
• /
• pp.29-38
• /
• 2002

The purpose of this paper was to examine early numerical representations between American and Korean children. Fifty-five first graders (35 Korean and 20 American) participated in the study. According to the findings of the current study, the author concluded that the Korean children had a stronger conception of base ten representations of numbers than that of the American children. The Korean children used various strategic reasoning such as decomposition and recomposition on the basis of base 10 structure to solve addition and subtraction problems effectively. However, the author cannot conclude that language differences would be the largest factor that would make Korean children sapient in the representations of base ten structures.

• Journal of the Korea Society of Computer and Information
• /
• v.5 no.4
• /
• pp.189-199
• /
• 2000

This Present study examines the Japanese Tourist basic law, its background as well as the formation process. Furthermore, this article also investigates the contents of the tourist basic law. In addition, I intend to accomplish a comparative study between a tourist policy enacted by the tourist policy council in 1995 years and the policy enacted in the 1960ㆍ70s. As the purpose of the study is concerned, the question related to a recomposition of contents of the Japanese tourist basic law in the 21st century is considered.

• The Journal of Society for e-Business Studies
• /
• v.12 no.4
• /
• pp.145-163
• /
• 2007

Service-Oriented Architecture is a style of information systems that enables the creation of applications that are built by combining loosely coupled and interoperable services. A service is an implementation of business functionality with proper granularity and invoked with well-defined interface. In service modeling, when the granularity of a service is finer, the reusability and flexibility of the service is lower. For solving this problem concerns with the service granularity, it is critical to identify and define coarse-grained services from the domain analysis model. In this paper, we define the process to identify services from the Use Case model elicited from domain analysis. A task tree is derived from Use Cases and their descriptions, and Use Cases are reconstructed by the composition and decomposition of the task tree. Reconstructed Use Cases are defined and specified as services. Because our method is based on the widely used UML Use Case models, it can be helpful to minimize time and cost for developing services in various platforms and domains.

• Journal of Biosystems Engineering
• /
• v.19 no.1
• /
• pp.42-49
• /
• 1994

Texturization of plant protein means the physical or chemical recomposition method of plant protein damaged during the extracting process of soybean oil. As a stable protein supplement, substituted for meat, needs of texturized products have been increased. Twin screw extruder is a very effective tool for texturization process as a physical method. This research, using defatted soy flour as raw material and twin screw extruder manufactured in domestic, showed that plant protein was texturized successfully on the operating conditions of barrel temperature of $120{\sim}140^{\circ}C$, material feed rate of 30~36kg/hr and water content of 20~25%. It also showed that the shape of die affected the texturization continuity.

• Journal of Industrial Technology
• /
• v.24 no.B
• /
• pp.29-36
• /
• 2004

The wind and wind-induced dynamic wind pressures fluctuate irregularly according to time and space. In this study, the proper orthogonal decomposition(POD) technique is applied to wind pressures acting on a tall tower model, and the following results are found: the along-wind and across-wind forces can be reconstructed by only four dominant POD modes, and the reconstructed errors are 4.71% and 22%, respectively for across-wind and along-wind directions. The physical meanings for dominant modes are also presented in the paper. The POD analysis can compress complex wind pressure data only by a few dominant modes and interpret spatio-temporal characteristics of wind pressure by novel way while existing statistical methods do not have such benefits.

• Korean Institute of Interior Design Journal
• /
• no.41
• /
• pp.266-274
• /
• 2003

Synectics is one of several techniques used to enhance brainstorming by taking a more active role and introducing metaphor and structure into the process. It is unclear at what level of specificity this should be formulated as a pattern. This thesis reviews recent computational as well as experimental work on analogical reasoning based on synectics. New results regarding information processing of analogical reasoning stages, major computational models and recent attempts to compare these models are reviewed. Computational models are also discussed in the computational as well as cognitive psychology perspectives. Future directions in analogical reasoning research are proposed. The following import is the need to accommodate the typology and normal assessment in the concrete circumstances where actual reasoning and problem solving take place. In order to get to this end, we used computational models by Thagard who take the stand of ‘Computational Philosophy of Science’, which assumes ‘Weak AI’ to explicate what constitute the very pecularity of Analogical Reasoning.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.52 no.9
• /
• pp.547-553
• /
• 2003

In this paper, we designed the PC-based analyzer of the power spectral density that could estimate the heart rate variability from time series data of R-R interval. The power spectral density estimated that it applied the autoregressive model to the measured electrocardiogram during a short period. Also, the characteristics of the designed analyzer are that it could process of the signal filtering, the generation and recomposition of time series and the feature extraction at the same time. Especially the analyzer reconstructed which applied the lowpass filter of the time series composed by the linear interpolation so as to enhance the signal-to-noise feature. We could estimate the power spectral density that confirmed a variety of power peak with low frequency range and high frequency rang of autonomic nerve by the heart rate variability.