• Journal of Digital Convergence
• /
• v.16 no.10
• /
• pp.205-212
• /
• 2018

With the development of high performance CPU / GPU, artificial intelligence algorithms such as deep neural networks, and a large amount of data, machine learning has been extended to various applications. In particular, a large amount of data collected from the Internet of Things, social network services, web pages, and public data is accelerating the use of machine learning. Learning data sets for machine learning exist in various formats according to application fields and data types, and thus it is difficult to effectively process data and apply them to machine learning. Therefore, this paper studied a method for building a learning data set for machine learning in accordance with standardized procedures. This paper first analyzes the requirement of learning data set according to problem types and data types. Based on the analysis, this paper presents the reference model to build learning data set for machine learning applications. This paper presents the target standardization organization and a standard development strategy for building learning data set.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.10
• /
• pp.116-123
• /
• 2013

In mobile healthcare applications, the reliable transmission of the bio-data is very important. In this article, we present a reliable bio-data transmission technique for mobile healthcare monitoring service at 6LoWPAN multi-hop wireless networks. In particular, we expand IEEE 11073-20601 protocol, and propose the reliable path construction for 6LoWPAN aimed to reliably provide mobile healthcare service over wireless sensor network, using IPv6 network. 6LoWPAN is recognized possibility because it is agree with sensor network by raising Adaptation layer on the MAC layer to transmit IPv6 packets. In this article proposed minimize the algorithm complexity and reliability routing protocol because the 6LoWPAN devices are suitable for low cost, small size and battery that can be used to health care system environment. And detailed procedures and algorithms are presented. We the proposed method to prove the superiority of using NS-3 for compareing with AODV protocol.

• Proceedings of the Korean Society of Near Infrared Spectroscopy Conference
• /
• 2001.06a
• /
• pp.1082-1082
• /
• 2001

Extrusion is one of the most important processes in polymer industry. The characterization of the polymer melt during processing will improve this process noticeably, One possibility of characterizing the actual processed polymer melt is the inline near infrared (NIR) spectroscopy, With this method several polymer properties can be observed during processing, e.g. composition, moisture ormechanical properties of the melt. For this purpose probes for transmission and reflection measurements have been developed, withstanding the high temperatures and pressures appearing during extrusion process (tested up to 300$^{\circ}C$ and 10 ㎫). For the transmission system an optical bypass was developed to eliminate disturbing spectral influences and hence increase the long term stability, which is the prerequisite for an industrial application. Measurements in transmission and reflection produced comparable results (or blending processes, where the prediction error was less than 1%. An optimum RMSEP of only 0.24% was found for preprocessed polymer blends measured in transmission on a laboratory extruder. A transflection measurement allowed for the first time the recording of relevant NIR-spectra in the screw area of an extruder. The application to a (PE+PP) blending process delivered promising results. This new measurement mode allows the observation of the ongoing processes within the screw area, which is of maximum Interest for reactive extrusion processes. Due to economic reasons the calibration transfer between different extrusion systems is also of high importance. Investigations on simulated and real-world spectra showed that a calibration transfer is possible. A new method alternatively to the well-known direct standardization procedures was developed, which is based on an automatic data pretreatment. This procedure delivers comparable results for the calibration transfer. Overall this paper presents concepts, components and algorithms for the inline near infrared (NIR) spectroscopy for polymer extrusion, which allows the use of it in a real industrial extrusion process.

• Journal of KIISE:Databases
• /
• v.34 no.4
• /
• pp.357-368
• /
• 2007

This paper proposes the SAMD(Synchronization Algorithms based on Message Digest) algorithm based on message digest in order to data synchronization between a database server and a mobile database. The SAMD algorithm makes the images at the database server and the mobile database using message digest tables and compares two images in order to select the rows needed to synchronization. If two images are different, the synchronization is progressed according to synchronization policy. The SAMD algorithm does not use techniques which are dependent the some database vendor and also triggers, stored procedures and timestamps. The SAMD only uses the standard SQL functions for the synchronization. Therefore the SAMD algorithm is used in any combinations of the database server and the mobile database because the SAMD algorithm is not depended on some database venders. Also the SAMD algorithm shows better performance compared with existing synchronization products.

• Journal of Digital Contents Society
• /
• v.9 no.3
• /
• pp.471-481
• /
• 2008

Test scheduling considering power dissipation is an effective technique to reduce the testing time of complex SoCs and to enhance fault coverage under limitation of allowed maximum power dissipation. In this paper, a modeling technique of test resources and a test scheduling algorithm for efficient test procedures are proposed and confirmed. For test resources modeling, two methods are described. One is to use the maximum point and next maximum point of power dissipation in test resources, the other one is to model test resources by partitioning of them. A novel heuristic test scheduling algorithm, using the extended-tree-growing-graph for generation of maximum embedded cores usable simultaneously by using relations between test resources and cores and power-dissipation-changing-graph for power optimization, is presented and compared with conventional algorithms to verify its efficiency.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2009.10a
• /
• pp.945-949
• /
• 2009

This paper proposes an enhanced motion estimation that is one of core parts affecting the coding performance and visual quality in video coding. Although the full search technique, which is the most basic method of the motion estimation, presents the best visual quality, its computational complexity is great, since the search procedures to find the best matched block with each block in the current frame are carried out for all points inside the search area. Thus, various fast algorithms to reduce the computational complexity and maintain good visual quality have been proposed. The PMVFAST adopted the MPEG-4 visual standard produces the visual quality near that by the full search technique with the reduced computational complexity. In this paper, we propose a new motion vector prediction method using median processing. The proposed method reduces the computational complexity for the motion estimation significantly. Experimental results show that the proposed algorithm is faster than the PMVFAST and better than the full search in terms of search speed and average PSNR, respectively.

• The KIPS Transactions:PartB
• /
• v.11B no.3
• /
• pp.353-360
• /
• 2004

The study on scene change detection in the compressed MPEG videos has been done by various approaches. However, most of these approacher accomplished scene change detection by carrying out decoding processes and then by comparing pixels with pixels. This approach it not suitable for real time applications owing to much computing time of decoding processes. Recently, the study on scene change detection algorithms using only information of compressed domain is becoming Increasingly important. In this paper, we propose a sequential access method as an efficient scene change detection algorithm in the compressed domain. According to the type of pictures in the compressed MPEG video streams (divided in I-blocks and each I-block into P-blocks), the proposed algorithm provides effective scene change detection by applying sequential access and block by block mechanism. The proposed sequential access method provides fast and accurate detection operation by reducing checking procedures of unnecessary pictures due to molt of operations in compressed domain and checking by block units. Also, this approach uses optimal algorithm to provide fast and accurate detection operation.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.40 no.10
• /
• pp.2035-2046
• /
• 2015

Recently, researches on resource allocation algorithms operating in a distributed way are widely conducted because of the increasing number of network nodes and the rapidly changing the network environment. In this paper, we propose Multi-Hop DESYNC(MH DESYNC), that is bio-inspired TDMA-based resource allocation scheme operating in a distributed manner in multi-hop networks. In this paper, we define a frame structure for the proposed MH DESYNC algorithm and firing message structure which is a reference for resource allocation and propose the related operating procedures. We show that MH DSYNC can resolve the hidden-node problem effectively and verify that each node shares resources fairly among its neighboring nodes. Through simulation evaluations, it is shown that MH DESYNC algorithm works well in a multi-hop networks. Furthermore, results show that MH DESYNC algorithm achieves better performance than CSMA/CA algorithm in terms of throughput.

• The KIPS Transactions:PartB
• /
• v.16B no.1
• /
• pp.47-54
• /
• 2009

Automatic object recognition in 3D measuring data is of great interest in many application fields e.g. computer vision, reverse engineering and digital factory. In this paper we present a software tool for a fully automatic object detection and parameter estimation in unordered and noisy point clouds with a large number of data points. The software consists of three interactive modules each for model selection, point segmentation and model fitting, in which the orthogonal distance fitting (ODF) plays an important role. The ODF algorithms estimate model parameters by minimizing the square sum of the shortest distances between model feature and measurement points. The local quadric surface fitted through ODF to a randomly touched small initial patch of the point cloud provides the necessary initial information for the overall procedures of model selection, point segmentation and model fitting. The performance of the presented software tool will be demonstrated by applying to point clouds.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.20 no.1
• /
• pp.9-18
• /
• 2007

This study shows an implementation of partial holes in an initial design domain in order to improve convergences of topology optimization algorithms. The method is associated with a bubble method as introduced by Eschenauer et al. to overcome slow convergence of boundary-based shape optimization methods. However, contrary to the bubble method, initial holes are only implemented for initializations of optimization algorithm in this approach, and there is no need to consider a characteristic function which defines hole's deposition during every optimization procedure. In addition, solid and void regions within the initial design domain are not fixed but merged or split during optimization Procedures. Since this phenomenon activates finite changes of design parameters without numerically calculating movements and positions of holes, convergences of topology optimization algorithm can be improved. In the present study, material topology optimization designs of Michell-type beam utilizing the initial design domain with initial holes of varied sizes and shapes is carried out by using SIMP like a density distribution method. Numerical examples demonstrate the efficiency and simplicity of the present method.