• The KIPS Transactions:PartB
• /
• v.10B no.6
• /
• pp.673-680
• /
• 2003

In this paper, we study the enhancement of VQ (Vector Quantization) design for text independent speaker recognition. In a concrete way, we present a non-iterative method which makes a vector quantization codebook and this method performs non-iterative learning so that the computational complexity is epochally reduced The proposed Classified Space VQ (CSVQ) design method for text Independent speaker recognition is generalized from Semi-noniterative VQ design method for text dependent speaker recognition. CSVQ contrasts with the existing desiEn method which uses the iterative learninE algorithm for every traininE speaker. The characteristics of a CSVQ design is as follows. First, the proposed method performs the non-iterative learning by using a Classified Space Codebook. Second, a quantization region of each speaker is equivalent for the quantization region of a Classified Space Codebook. And the quantization point of each speaker is the optimal point for the statistical distribution of each speaker in a quantization region of a Classified Space Codebook. Third, Classified Space Codebook (CSC) is constructed through Sample Vector Formation Method (CSVQ1, 2) and Hyper-Lattice Formation Method (CSVQ 3). In the numerical experiment, we use the 12th met-cepstrum feature vectors of 10 speakers and compare it with the existing method, changing the codebook size from 16 to 128 for each Classified Space Codebook. The recognition rate of the proposed method is 100% for CSVQ1, 2. It is equal to the recognition rate of the existing method. Therefore the proposed CSVQ design method is, reducing computational complexity and maintaining the recognition rate, new alternative proposal and CSVQ with CSC can be applied to a general purpose recognition.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.48 no.4
• /
• pp.285-293
• /
• 2020

In this paper, to improve Synthetic Aperture Radar (SAR) image quality, the effect of time synchronization error in the EGI/IMU (Embedded GPS/INS, Inertial Measurement Unit) integrated system is analyzed and state augmentation is applied to compensate it. EGI/IMU integrated system is widely used as a SAR motion measurement algorithm, which consists of EGI mounted to obtain the trajectory and IMU mounted on the SAR antenna. In an EGI/IMU integrated system, a time synchronization error occurs when the clocks of the sensors are not synchronized. Analysis of the effect of time synchronization error on navigation solutions and SAR images confirmed that the time synchronization error deteriorates SAR image quality. The state augmentation is applied to compensate for this and as a result, the SAR image quality does not decrease. In addition, by analyzing the performance and the observability of the time synchronization error according to the maneuver, it was confirmed that the time-variant maneuver such as rotational motion is necessary to estimate the time synchronization error adequately. In order to reduce the influence of the time synchronization error on the SAR image, the time synchronization error must be compensated by performing maneuver changing over time such as a rotation before SAR operation.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.3 no.2
• /
• pp.87-96
• /
• 1999

To analyze the dynamic behavior of structure, direct integration and mode superposition may be utilized in time domain analysis. As finite number of frequencies can give relatively exact solutions, mode superposition is preferable in analyzing structural behavior. In non-linear analysis, however, mode superposition is seldom used since time-varying element stiffness changes stiffness matrix, and the change of stiffness matrix leads to the change of essential constants - natural frequencies and mode shapes. In spite of these difficulties, there are some attempts to adopt mode superposition because of low cost compared to direct integration, but the result is not satisfactory. In this paper, a method using mode superposition in non-linear analysis is presented by separating local element stiffness from global stiffness matrix with the difference between linear and non-linear restoring forces to the external force vectors included. Moreover, the hysteresis model changing with the relative deformation in each floor makes it possible to analyze non-linear behavior of structure. The proposed algorithm is applied to shear beam model and the maximum displacement is compared with the result using direct integration method.

• Journal of Distribution Science
• /
• v.15 no.6
• /
• pp.77-96
• /
• 2017

Purpose - This study aims to develop correspondence strategies to the environment change in domestic retail store types. Recently, new types of retails have emerged in retail industries. Therefore, trade area platform has developed focusing on the speed of data, no longer trade area from district border. Besides, 'trade area smart' brings about change in retail types with the development of giga internet. Thus, context shopping is changing the way of consumers' purchase pattern through data capture, technology capability, and algorithm development. For these reasons, the sales estimation model has been shown to be flawed using the notion of former scale and time, and it is necessary to construct a new model. Research design, data, and methodology - This study focuses on measuring retail change in large multi-shopping mall for the outlook for retail industry and competition for trade area with the theoretical background understanding of retail store types and overall domestic retail conditions. The competition among retail store types are strong, whereas the borders among them are fading. There is a greater need to analyze on a new model because sales expectation can be hard to get with business area competition. For comprehensive research, therefore, the research method based on the statistical analysis was excluded, and field survey and literature investigation method were used to identify problems and propose an alternative. In research material, research fidelity has improved with complementing research data related with retail specialists' as well as department stores. Results - This study analyzed trade area survival and its pattern through sales estimation and empirical studies on trade areas. The sales estimation, based on Huff model system, counts the number of households shopping absorption expectation from trade areas. Based on the results, this paper estimated sales scale, and then deducted modified probability model. Conclusions - In times of retail store chain destruction and off-line store reorganization, modified Huff model has problems in estimating sales. Transformation probability model, supplemented by the existing problems, was analyzed to be more effective in competitiveness business condition. This study offers a viable alternative to figure out related trade areas' sale estimation by reconstructing new-modified probability model. As a result, the future task is to enlarge the borders from IT infrastructure with data and evidence based business into DT infrastructure.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.1
• /
• pp.118-125
• /
• 2016

This study evaluates the performance of image to patient registration algorithm by using stereo vision and CT image for facial region surgical navigation. For the process of image to patient registration, feature extraction and 3D coordinate calculation are conducted, and then 3D CT image to 3D coordinate registration is conducted. Of the five combinations that can be generated by using three facial feature extraction methods and three registration methods on stereo vision image, this study evaluates the one with the highest registration accuracy. In addition, image to patient registration accuracy was compared by changing the facial rotation angle. As a result of the experiment, it turned out that when the facial rotation angle is within 20 degrees, registration using Active Appearance Model and Pseudo Inverse Matching has the highest accuracy, and when the facial rotation angle is over 20 degrees, registration using Speeded Up Robust Features and Iterative Closest Point has the highest accuracy. These results indicate that, Active Appearance Model and Pseudo Inverse Matching methods should be used in order to reduce registration error when the facial rotation angle is within 20 degrees, and Speeded Up Robust Features and Iterative Closest Point methods should be used when the facial rotation angle is over 20 degrees.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.44 no.1
• /
• pp.146-154
• /
• 2007

In this paper, we have presented a real-time transfer mechanism for the delay-sensitive data in WSNs (Wireless Sensor Networks). The existing methods for real-time data transfer select a path whose latency is shortest or the number of hops is least. Although the approaches of these methods are acceptable, they do not always work as efficiently as they can because they had no consideration for the link error rates. In the case of transmission failures on links, they can not guarantee the end-to-end real-time transfer due to retransmissions. Therefore, we have proposed an algorithm to select a real-time transfer path in consideration of the link error rates. Our mechanism estimates the 1-hop delay based on the link error rate between two neighboring nodes, which in turn enables the calculation of the expected end-to-end delay. A source node comes to choose a path with the shortest end-to-end delay as a real-time route, and sends data along the path chosen. We performed various experiments changing the link error rates and discovered that this proposed mechanism improves the speed of event-to-sink data transfer and reduces delay jitter. We also found that this mechanism prevents additional energy consumption and prolongs network lifetime, resulting from the elative reduction of transmission failures and retransmissions.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.44 no.4 s.316
• /
• pp.42-47
• /
• 2007

To reduce the chattering errors of reed switch sensors in the automatic remote measurement of water meter a reed switch sensor was analyzed and improved. The operation of reed switch sensors can be described as a mechanical contact switch by approximation of permanent magnet piece to generate an electrical pulse. The reed switch sensors are used mostly in measurement application to detect the rotational or translational displacement. To apply for water flow measurement devices, the reed switch sensors should keep high reliability. They are applied for the electronic digital type of water flow meters. The reed switch sensor is just mounted simply on the conventional mechanical type flow meter. A small magnet is attached on a pointer of the water meter counter rotor. Inside the reed sensor two steel leaf springs make mechanical contact and apart repeatedly as rotation of flow meter counter. The counting electrical contact pulses can be converted as the water flow amount. The MCU sends the digital flow rate data to the server using the wireless communication network. But the digital data is occurred difference or won by chattering noise. The reed switch sensor contains chattering error by it self at the force equivalent position. The vibrations such as passing vehicle near to the switch sensor installed location causes chattering. In order to reduce chattering error, most system uses just software methods, for example using filter algorithm and also statistical calibration methods. The chattering errors were reduced by changing leaf spring structure using mechanical characteristics.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.57 no.12
• /
• pp.2351-2357
• /
• 2008

Noninvasive radial artery pulse wave has been widely used not only for the pulse wave analysis(PWA) itself but also for assessment of arterial stiffness with estimated aortic pulse wave from peripheral pulse wave. However, it has been found that the deformation of pulse shape can be caused readily by changing measuring position, indentation pressure, and so on. So, in this study, we have developed a system which can measure radial pulse wave and skin displacement simultaneously while the indentation body goes down to occlude subject's radial artery. This system can be divided into a measuring apparatus part, an indentation control hardware part, a data acquisition part and a control and computation part. And, the measuring apparatus consists of an arm-rest, a step motor, an indentation body, a laser displacement sensor(LK-G30, Keyence Co.) and pulse wave sensor. Under load-free condition and radial artery loaded condition, the evaluation of developed system has been performed. From these results, we can conclude: 1) The developed system can control the indentation body quantitatively and the adopted laser displacement sensor shows linear output characteristic even with skin as a reflector. 2) This system can measure the pulse wave and the displacement of indentation body, that is, skin displacement simultaneously at each specific level of indentation body. 3) This system can provide the number of motor steps used to get down the indentation body, the measured skin displacement, the calculated indentation pressure, the calculated pulse pressure and the pulse waveform as well as the information generated by combining these with each others. 4) This system can reveal the relationship between the morphological changes of pulse wave and the estimated displacement of radial artery wall by indentation. Consequently, the developed system can furnish more abundant information on radial artery than previous diagnosis systems based on tonometric measurement. In further study, we expect to setup the standard measuring process and to concrete the algorithm for the estimation of radial artery's diameter and of displacement of radial artery's wall. Furthermore, with well designed clinical studies, we hope to turn out the usefulness of developed system in the field of cardiovascular system evaluation.

• The KIPS Transactions:PartC
• /
• v.19C no.2
• /
• pp.135-144
• /
• 2012

All the servers in a traditional server cluster environment are kept On. If the request load reaches to the maximum, we exploit its maximum possible performance, otherwise, we exploit only some portion of maximum possible performance so that the efficiency of server power consumption becomes low. We can improve the efficiency of power consumption by controlling power mode of servers according to load situation, that is, by making On only minimum number of servers needed to handle current load while making Off the remaining servers. In the existing power mode control method, they used a static policy to decide server power mode at a fixed time interval so that it cannot adapt well to the dynamically changing load situation. In order to improve the existing method, we propose a dynamic server power control algorithm. In the proposed method, we keep the history of server power consumption and, based on it, predict whether power consumption increases in the near future. Based on this prediction, we dynamically change the time interval to decide server power mode. We performed experiments with a cluster of 30 PCs. Experimental results show that our proposed method keeps the same performance while reducing 29% of power consumption compared to the existing method. In addition, our proposed method allows to increase the average CPU utilization by 66%.

• Convergence Security Journal
• /
• v.11 no.1
• /
• pp.71-77
• /
• 2011

NCW(Network Centric- Warfare) encompasses the concept to use computer data processing and network linkage communications techniques, share information and furthermore, enhance the effectiveness of computer-operating systems. As IT(Information & Technology) have become developed in the recent years, the existing warfare system-centered conventional protocol is not use any longer. Instead, network-based NCW is being widely-available, today. Under this changing computer environment, it becomes important to establish algorithm and build the stable communication systems. Tools to identify malign node factors through Wireless Ad-hoc network cause a tremendous error to analyze and use paths of even benign node factors misreported to prove false without testing or indentifying such factors to an adequate level. These things can become an obstacle in the process of creating the optimum network distribution environment. In this regard, this thesis is designed to test and identify paths of benign node factors and then, present techniques to transmit data through the most significant open short path, with the tool of MP-SAR Protocol, security path search provider, in Ad-hoc NCW environment. Such techniques functions to identify and test unnecessary paths of node factors, and thus, such technique users can give an easy access to benign paths of node factors.