• Journal of Korea Multimedia Society
• /
• v.10 no.9
• /
• pp.1220-1228
• /
• 2007

This paper presents a new system for online signature verification, approaching for finding gaps between a point-to-point matching and a segment-to-segment matching. Each matching algorithm has been separately used in previous studies. Various features with respect to each matching algorithm have been extracted for solving two-class classification problem. We combined advantages of the two algorithms to implement an efficient system for online signature verification. In the proposed method, extreme feints are used to extract writer-dependent features. In addition, using the writer-dependent features proves to be more adaptive than using writer-independent features in terms of efficiency of classification and verification in this paper.

• KIPS Transactions on Computer and Communication Systems
• /
• v.11 no.12
• /
• pp.445-452
• /
• 2022

Class distribution of unbalanced data is an important part of the digital world and is a significant part of cybersecurity. Abnormal activity of unbalanced data should be found and problems solved. Although a system capable of tracking patterns in all transactions is needed, machine learning with disproportionate data, which typically has abnormal patterns, can ignore and degrade performance for minority layers, and predictive models can be inaccurately biased. In this paper, we predict target variables and improve accuracy by combining estimates using Synthetic Minority Oversampling Technique (SMOTE) and Light GBM algorithms as an approach to address unbalanced datasets. Experimental results were compared with logistic regression, decision tree, KNN, Random Forest, and XGBoost algorithms. The performance was similar in accuracy and reproduction rate, but in precision, two algorithms performed at Random Forest 80.76% and Light GBM 97.16%, and in F1-score, Random Forest 84.67% and Light GBM 91.96%. As a result of this experiment, it was confirmed that Light GBM's performance was similar without deviation or improved by up to 16% compared to five algorithms.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.5 no.1
• /
• pp.89-101
• /
• 2001

The conventional pattern-methods for discrimination of acupoint, meridian line which is the basic object of diagnosis and medical treatment in oriental medicine is discriminated the conduction point by the stimulation in body skin with DC. But, it is not sufficient to truth in discrimination ratio, coincident ratio, body effect, reproductivity. Therefore, this paper is extracting the optimal parameter of frequency and waveform in order to improve the conventional pattern, and proposing the SPAC(Single Power Alternative Current) stimulus pattern applying that. Also, this algorithm proposes to be able to discriminate with low pressure of the electrode by displaying in the level meter both the absolution and relation value of the skin current. It is able to decrease pain and body effect by electrode pressure and discriminate acupoint regardless of skin current in difficult discrimination spot. It is compared the performance of system applying the extracted optimal parameter and algorithm, and it is confirmed that there is difference in discrimination parameter of acupoint reacted to the individual and the meridian. It is compared that discrimination, coincident ratio of the traditional acupoints as the acupoint stimulation pattern. It is confirmed truth of optimal parameter and discrimination algorithm. Keyword: Meridian, Discrimination, Coincident, Body effect, Reproductivity, SPAC, Optimal parameter.

• Journal of Convergence for Information Technology
• /
• v.11 no.6
• /
• pp.219-225
• /
• 2021

As for the battery package method, a prismatic package method is preferred for stability reasons, but it is rapidly expanding due to the stability verification of a pouch type package. The pouch type using the lamination process has an advantage of high battery energy density because it can reduce space waste, but has a disadvantage of low productivity. Therefore, in this paper, by extracting edge detection algorithm precision, pattern algorithm precision, and motion controller recall rate by improving backlight lighting fixtures to minimize light diffusion, securing standards for stereo camera position relationship displacement monitoring, and securing standards for lens release monitoring. We propose to implement a system that ensures accuracy and reliability in positioning. As a result of the experiment, the proposed system shows an average error range of 0.032mm for edge detection, 0.02mm for pattern algorithm, and 0.014mm for motion controller, thus ensuring the accuracy and reliability of the positioning mechanism.

• Journal of the Korea Convergence Society
• /
• v.9 no.12
• /
• pp.33-38
• /
• 2018

Recently, we are trying to predict the emotion from various information and to convey the emotion information that the supervisor wants to inform the audience. In addition, audiences intend to understand the flow of emotions through various information of non-dialogue parts, such as cinematography, scene background, background sound and so on. In this paper, we propose to extract emotions by mixing not only the context of scripts but also the cinematography information such as color, background sound, composition, arrangement and so on. In other words, we propose an emotional prediction system that learns and distinguishes various emotional expression techniques into dialogue and non-dialogue regions, contributes to the completeness of the movie, and quickly applies them to new changes. The precision of the proposed system is improved by about 5.1% and 0.4%, and the recall is improved by about 4.3% and 1.6%, respectively, when compared with the modified n-gram and morphological analysis.

• Journal of Broadcast Engineering
• /
• v.23 no.6
• /
• pp.780-789
• /
• 2018

Nowadays, gun crimes occur very frequently not only in public places but in alleyways around the world. In particular, it is essential to detect a person armed by a pistol to prevent those crimes since small guns, such as pistols, are often used for those crimes. Because conventional works for armed person detection have treated an armed person as a single object in an input image, their accuracy is very low. The reason for the low accuracy comes from the fact that the gunman is treated as a single object although the pistol is a relatively much smaller object than the person. To solve this problem, we propose a novel algorithm called APDA(Armed Person Detection Algorithm). APDA detects the armed person using in a post-processing the positions of both wrists and the pistol achieved by the CNN-based human body feature detection model and the pistol detection model, respectively. We show that APDA can provide both 46.3% better recall and 14.04% better precision than SSD-MobileNet.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2008.06a
• /
• pp.181-181
• /
• 2008

DLC (Diamond-like Carbon) 박막은 높은 내마모성과 낮은 마찰 계수, 화학적 안정성 및 적외선 영역에서의 높은 투과율과 낮은 광 반사도, 높은 전기저항과 낮은 유전율, 전계방출특성 등 여러 가지 장점을 가진 물질이다[1]. 최근에는 DLC 박막의 여러 장점들과 산과 염기 유기용매에 대한 화학적 안정성으로 인하여 인조관절에서 인공심장의 판막에 이르기까지 의공학 관련 부품소재로 응용되고 있으며 내구성과 안정성에 있어서 탁월한 성능을 보여주고 있다. 또한 DLC 박막의 높은 경도와 낮은 마찰 계수, 부드러운 박막 표면 (수nm의 RMS 거칠기)의 장점을 살려 마그네틱 미디어와 하드디스크의 슬라이딩 표면에 사용되어지고, MEMS (Micro-Electro Mechanical System) 소자와 MMAs (Moving Mechanical Assemblies)의 고체윤활코팅으로 활용하여 미세기계의 내구성과 성능 향상을 도모할 수 있다. 이와 같이 DLC 박막은 다양한 분야에 응용되고 있으며, 박막이 지닌 여러 가지 장점들로 인하여 더 많은 분야에 응용될 가능성을 지닌 물질이다. 그러나 수 ${\mu}m$이상의 두께에서 박막이 높은 잔류응력 (residual stress)을 가지고, 열에 취약하여 이의 개선에 관한 연구들이 진행되어 지고 있다 [2]. 따라서 사용되는 목적에 따라 용도에 맞는 양질의 DLC 박막을 합성하기 위해선 합성 장치의 개발과 다양한 실험을 통한 최적의 합성조건 도출 등의 노력이 요구된다. 또한 DLC 박막 합성시의 여러 가지 증착 방법에 따른 박막 물성에 대한 재현성 확보 및 박막 증착에 관한 명확한 메커니즘 규명이 아직까지는 불분명하여 이에 관한 연구가 시급하다. 따라서 본 연구에서는 MEMS 소자와 MMAs의 고체윤활코팅으로 사용가능한 DLC 박막을 RF PECVD (Plasma Enhanced Vapor Deposition) 방식으로 합성하고 후열처리 온도에 따른 DLC 박막의 마찰계수 변화를 박막에 훼손을 주지 않는 FFM (Friction Force Microscopy) 방식을 사용하여 분석하였다.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.11
• /
• pp.15-22
• /
• 2016

In satellite communication, transmission performance is degraded due to long propagation delay and relatively high data loss compared to terrestrial network. In this paper, We propose Hybrid TCP PEP scheme with XOR coding and Hybla TCP, which reduces the transmission performance degradation due to the transmission delay time. Experimental results show that the proposed method improves the file transfer rate by more than 10% in the environment with high channel error rate. Therefore, Hybrid TCP, which is a mixture of XOR coding method and TCP Hybla, is considered to contribute to the improvement of transmission speed in satellite communication when applied to connection split PEP.

• Smart Media Journal
• /
• v.7 no.4
• /
• pp.17-23
• /
• 2018

We proposed a LSTM-based RNN model that can effectively perform the automatic spacing characteristics. For those long or noisy sentences which are known to be difficult to handle within Neural Network Learning, we defined a proper input data format and decoding data format, and added dropout, bidirectional multi-layer LSTM, layer normalization, and attention mechanism to improve the performance. Despite of the fact that Sejong corpus contains some spacing errors, a noise-robust learning model developed in this study with no overfitting through a dropout method helped training and returned meaningful results of Korean word spacing and its patterns. The experimental results showed that the performance of LSTM sequence-to-sequence model is 0.94 in F1-measure, which is better than the rule-based deep-learning method of GRU-CRF.

• Korean Journal of Optics and Photonics
• /
• v.15 no.5
• /
• pp.423-428
• /
• 2004

The development of a compact and high performance MWIR thermal imaging sensor based on the SOFRADIR 320${\times}$240 element IRCCD detector is described. The sensor has 20 magnification zoom optics with the maximum 40$^{\circ}$${\times}$30$^{\circ}$ of super wide field of view and 7.6 cycles/mrad of resolving power with the operation of attached micro-scanning system. In order to correct nonuniformities of detector arrays, we have proposed a multi-point correction method using defocusing of the optics and we have acquired the highest quality images. The MRTD of our system shows good results below 0.05K at spatial frequency 1 cycles/mrad at narrow field of view. Experimental data and obtained performances are presented and discussed.