• Journal of the Korea Society of Computer and Information
• /
• v.29 no.5
• /
• pp.93-100
• /
• 2024

With the advent of the Fourth Industrial Revolution, cutting-edge technologies such as artificial intelligence, big data, the Internet of Things, and cloud computing are driving innovation across industries. These technologies are generating massive amounts of data that many companies are leveraging. However, there is a notable reluctance among users to share sensitive information due to the privacy risks associated with collecting personal data. This is particularly evident in the healthcare sector, where the collection of sensitive information such as patients' medical conditions poses significant challenges, with privacy concerns hindering data collection and analysis. This research presents a novel technique for collecting and analyzing medical data that not only preserves privacy, but also effectively extracts statistical information. This method goes beyond basic data collection by incorporating a strategy to efficiently mine statistical data while maintaining privacy. Performance evaluations using real-world data have shown that the propose technique outperforms existing methods in extracting meaningful statistical insights.

• The Journal of the Convergence on Culture Technology
• /
• v.10 no.1
• /
• pp.471-476
• /
• 2024

With the spread of various smart devices and computing devices, big data generation is occurring widely. Machine learning is an algorithm that performs reasoning by learning data patterns. Among the various machine learning algorithms, the algorithm that attracts attention is deep learning based on neural networks. Deep learning is achieving rapid performance improvement with the release of various applications. Recently, among deep learning algorithms, attempts to analyze data using graph structures are increasing. In this study, we present a graph generation method for transferring to a deep learning network. This paper proposes a method of generalizing node properties and edge weights in the graph generation process and converting them into a structure for deep learning input by presenting a matricization We present a method of applying a linear transformation matrix that can preserve attribute and weight information in the graph generation process. Finally, we present a deep learning input structure of a general graph and present an approach for performance analysis.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.4
• /
• pp.77-82
• /
• 2024

Recent advancements in artificial intelligence (AI) technology have led to an increase in the implementation of AI applications in mobile environments. However, due to the limited resources in mobile devices compared to desktops and servers, there is growing interest in research aimed at efficiently executing AI workloads on mobile platforms. While most studies focus on offloading to edge or cloud solutions to mitigate computing resource constraints, research on the characteristics of file I/O related to storage access in mobile settings remains underexplored. This paper analyzes file I/O traces generated during the execution of deep learning applications in mobile environments and investigates how they differ from traditional mobile workloads. We anticipate that the findings of this study will be utilized to design future smartphone system software more efficiently, considering the file access characteristics of deep learning.

• Journal of Internet Computing and Services
• /
• v.4 no.3
• /
• pp.57-67
• /
• 2003

Software product lines are to attain the rapid development of qualify applications by concretizing the general components populated in software assets and assembling them according to the predefined architectures. For supporting the construction of the software product lines, this paper proposes a component code generation techniques based on GenVoca architecture and XML/XSLT technologies, In addition, CompGenX(Component Generator using XML), a component code generation system, is proposed on the basis of this techniques. By providing reconfigurability of component at the time of code generation, CompGenX allows the reusers to create the component source code that is appropriate to their purpose, In this system, the process of the component development is divided into two tasks which are the component family construction task and the component reuse task, For the component family construction, CompGenX provides the feature modeling tool for domain analysis and the domain architecture definition tool. Also, it provides the tool for building the component configuration know1edge specification and the code templates, For the component reuse task, it offers the component family search tool. the component customizing tool and the component code generator. Component code generation techniques and system in this paper should be applicable as basic technology to build the component-based software product lines.

• Journal of Internet Computing and Services
• /
• v.19 no.5
• /
• pp.43-54
• /
• 2018

Images taken in haze weather are characteristic of low contrast and poor visibility. The process of reconstructing clear-weather image from a hazy image is called dehazing. The main challenge of image dehazing is to estimate the transmission map or depth map for an input hazy image. In this paper, we propose a single image dehazing method by utilizing the Generative Adversarial Network(GAN) for accurate depth map estimation. The proposed GAN model is trained to learn a nonlinear mapping between the input hazy image and corresponding depth map. With the trained model, first the depth map of the input hazy image is estimated and used to compute the transmission map. Then a guided filter is utilized to preserve the important edge information of the hazy image, thus obtaining a refined transmission map. Finally, the haze-free image is recovered via atmospheric scattering model. Although the proposed GAN model is trained on synthetic indoor images, it can be applied to real hazy images. The experimental results demonstrate that the proposed method achieves superior dehazing results against the state-of-the-art algorithms on both the real hazy images and the synthetic hazy images, in terms of quantitative performance and visual performance.

• Journal of Radiation Protection and Research
• /
• v.17 no.1
• /
• pp.43-56
• /
• 1992

As one of the methods to ameliorate the ray effects which are the nature of anomalous computational effects due to the discretization of the angular variable in discrete ordinates approximations, a computational program, named TWODET (TWO dimensional Discrete Element Transport), has developed in 2 dimensional cartesian coordinates system using the discrete elements method, in which the discrete angle quadratures are steered by the spatially dependent angular fluxes. The results of the TWODET calculation with K-2, L-3 discrete angular quadratures, in the problem of a centrally located, isotropically emitting flat source in an absorbing square, are shown to be more accurate than that of the DOT 4.3 calculation with S-10 full symmetry angular quadratures, in remedy of the ray effect at the edge flux distributions of the square. But the computing time of the TWODET is about 4 times more than that of the DOT 4.3. In the problem of vacuum boundaries just outside of the source region in an absorbing square, the results of the TWODET calculation are shown severely anomalous ray effects, due to the sudden discontinuity between the source and the vacuum, like as the results of the DOT 4.3 calculation. In the probelm of an external source in an absorbing square in which a highly absorbing medium is added, the results of the TWODET calculation with K-3, L-4 show a good ones like as, somewhat more than, that of the DOT 4.3 calculation with S-10.

• Journal of KIISE:Computing Practices and Letters
• /
• v.10 no.6
• /
• pp.514-528
• /
• 2004

An emotion-based video scene retrieval algorithm is proposed in this paper. First, abrupt/gradual shot boundaries are detected in the video clip representing a specific story Then, five video features such as 'average color histogram' 'average brightness', 'average edge histogram', 'average shot duration', and 'gradual change rate' are extracted from each of the videos and mapping between these features and the emotional space that user has in mind is achieved by an interactive genetic algorithm. Once the proposed algorithm has selected videos that contain the corresponding emotion from initial population of videos, feature vectors from the selected videos are regarded as chromosomes and a genetic crossover is applied over them. Next, new chromosomes after crossover and feature vectors in the database videos are compared based on the similarity function to obtain the most similar videos as solutions of the next generation. By iterating above procedures, new population of videos that user has in mind are retrieved. In order to show the validity of the proposed method, six example categories such as 'action', 'excitement', 'suspense', 'quietness', 'relaxation', 'happiness' are used as emotions for experiments. Over 300 commercial videos, retrieval results show 70% effectiveness in average.

• Journal of Internet Computing and Services
• /
• v.7 no.2
• /
• pp.23-35
• /
• 2006

This work presents o novel method which extract facial region und features from motion picture automatically and controls the 3D facial expression in real time. To txtract facial region and facial feature points from each color frame of motion pictures a new nonparametric skin color model is proposed rather than using parametric skin color model. Conventionally used parametric skin color models, which presents facial distribution as gaussian-type, have lack of robustness for varying lighting conditions. Thus it needs additional work to extract exact facial region from face images. To resolve the limitation of current skin color model, we exploit the Hue-Tint chrominance components and represent the skin chrominance distribution as a linear function, which can reduce error for detecting facial region. Moreover, the minimal facial feature positions detected by the proposed skin model are adjusted by using edge information of the detected facial region along with the proportions of the face. To produce the realistic facial expression, we adopt Water's linear muscle model and apply the extended version of Water's muscles to variation of the facial features of the 3D face. The experiments show that the proposed approach efficiently detects facial feature points and naturally controls the facial expression of the 3D face model.

• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.4
• /
• pp.284-288
• /
• 2009

With explosively growing PPI databases, the computational approach for a prediction and configuration of PPI network has been a big stream in the bioinformatics area. Recent researches gradually consider physicochemical properties of proteins and support high resolution results with integration of experimental results. With regard to current research trend, it is very close future to complete a PPI network configuration of each organism. However, direct applying the PPI network to real field is complicated problem because PPI network is only a set of co-expressive proteins or gene products, and its network link means simple physical binding rather than in-depth knowledge of biological process. In this paper, we suggest a protein functional flow model which is a directed network based on a protein functions' relation of signaling transduction pathway. The vertex of the suggested model is a molecular function annotated by gene ontology, and the relations among the vertex are considered as edges. Thus, it is easy to trace a specific function's transition, and it can be a constraint to extract a meaningful sub-path from whole PPI network. To evaluate the model, 11 functional flow models of Homo sapiens were built from KEGG, and Cronbach's alpha values were measured (alpha=0.67). Among 1023 functional flows, 765 functional flows showed 0.6 or higher alpha values.

• KIPS Transactions on Computer and Communication Systems
• /
• v.11 no.10
• /
• pp.323-332
• /
• 2022

As various services using AI technology are being developed, much attention is being paid to AI service production. Recently, AI technology is acknowledged as one of ICT services, a lot of research is being conducted for general-purpose AI service production. In this paper, I describe the research results in terms of systems for AI service production, focusing on the distribution and production of machine learning models, which are the final steps of general machine learning development procedures. Three different Ubuntu systems were built, and experiments were conducted on the system, using data from 2017 validation COCO dataset in combination of different AI models (RFCN, SSD-Mobilenet) and different communication methods (gRPC, REST) to request and perform AI services through Tensorflow serving. Through various experiments, it was found that the type of AI model has a greater influence on AI service inference time than AI machine communication method, and in the case of object detection AI service, the number and complexity of objects in the image are more affected than the file size of the image to be detected. In addition, it was confirmed that if the AI service is performed remotely rather than locally, even if it is a machine with good performance, it takes more time to infer the AI service than if it is performed locally. Through the results of this study, it is expected that system design suitable for service goals, AI model development, and efficient AI service production will be possible.