• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.43 no.3 s.309
• /
• pp.30-38
• /
• 2006

In this paper, we improved to drops recognition rate according to distance using distance and depth information with 3D from stereo face images. A monocular face image has problem to drops recognition rate by uncertainty information such as distance of an object, size, moving, rotation, and depth. Also, if image information was not acquired such as rotation, illumination, and pose change for recognition, it has a very many fault. So, we wish to solve such problem. Proposed method consists of an eyes detection algorithm, analysis a pose of face, md principal component analysis (PCA). We also convert the YCbCr space from the RGB for detect with fast face in a limited region. We create multi-layered relative intensity map in face candidate region and decide whether it is face from facial geometry. It can acquire the depth information of distance, eyes, and mouth in stereo face images. Proposed method detects face according to scale, moving, and rotation by using distance and depth. We train by using PCA the detected left face and estimated direction difference. Simulation results with face recognition rate of 95.83% (100cm) in the front and 98.3% with the pose change were obtained successfully. Therefore, proposed method can be used to obtain high recognition rate with an appropriate scaling and pose change according to the distance.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.5
• /
• pp.21-27
• /
• 1999

In this paper, we propose an improved MLP-based English pronunciation dictionary generator to apply to the variable vocabulary word recognizer. The variable vocabulary word recognizer can process any words specified in Korean word lexicon dynamically determined according to the current recognition task. To extend the ability of the system to task for English words, it is necessary to build a pronunciation dictionary generator to be able to process words not included in a predefined lexicon, such as proper nouns. In order to build the English pronunciation dictionary generator, we use context-dependent grapheme-to-phoneme multi-layer perceptron(MLP) architecture for each grapheme. To train each MLP, it is necessary to obtain grapheme-to-phoneme training data from general pronunciation dictionary. To automate the process, we use dynamic programming(DP) algorithm with some distance metrics. For training and testing the grapheme-to-phoneme MLPs, we use general English pronunciation dictionary with about 110 thousand words. With 26 MLPs each having 30 to 50 hidden nodes and the exception grapheme lexicon, we obtained the word accuracy of 72.8% for the 110 thousand words superior to rule-based method showing the word accuracy of 24.0%.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.49 no.3
• /
• pp.38-46
• /
• 2012

Atrial fibrillation is needed to be detected at paroxysmal stage and to be treated. But, paroxysmal atrial fibrillation ECG is hardly obtained with 12-lead electrocardiographs but Holter systems. Presently, the averaged beat subtraction(ABS) method is solely used to estimate atrial fibrillatory waves even with somewhat large residual error. As an alternative, in this study, we suggested an ESAF(event-synchronous adaptive filter) based algorithm, in which the AF ECG was treated as a primary input and event-synchronous impulse train(ESIT) as a reference. And, ESIT was generated so to be synchronized with the ventricular activity by detecting QRS complex. We tested proposed algorithm with simulated AF ECGs and real AF ECGs. As results, even with low computational cost, this ESAF based algorithm showed better performance than the ABS method and comparable performance to algorithm based on PCA(principal component analysis) or SVD(singular value decomposition). We also proposed an expanded version of ESAF for some AF ECGs with multi-morphologic ventricular activities and this also showed reasonable performance. Ultimately, with Holter systems including our proposed algorithm, atrial activity signal can be precisely estimated in real-time so that it will be possible to calculate atrial fibrillatory rate and to evaluate the effect of anti-arrhythmic drugs.

• Spatial Information Research
• /
• v.10 no.1
• /
• pp.1-14
• /
• 2002

Due to the rapid development in information technology, countries of advanced information technology and multi-national information communication industries have expanding their invesment in constructing a cyber territory or a cyber city, and even a cyber globe that combines the real world with the cyber world. As such, in order to keep up with the competition with other countries to secure the Invisible Continent, Korea cannot afford to fall behind in the preparation and efforts to construct a cyber territory. Therefore in preparation for the inevitable establishment of a cyber territory, this study defined the concept of a cyber territory clearly and proposed a promotional strategy needed by the government for the establishment of a cyber territory. A cyber territory is 'the dynamic second territory that realizes various values in various aspects through organically combining innumerable active bodies in a surreal space and that are created by systematically and optimally connecting the physical topographical space and activity of the first nation to the cyberspace'. To explain further, a cyber territory is defined as another simulated space not only to manage the land systematically and deal with administrative services far the greater population, but also to contain economic activities of corporations and the citizens' everyday lives in a virtual reality by digitizing the entire territory including even the sea. In order to establish such cyber territory, it is necessary to revise related laws and policies, to foster related technology and industry as a main engine for national development, to promote public awareness, and to train related human resources.

• Asian Pacific Journal of Cancer Prevention
• /
• v.15 no.14
• /
• pp.5773-5779
• /
• 2014

Background: Since cervical cancer can be prevented and controlled through human papillomavirus (HPV) vaccination, it is important to train health care providers and provide them with appropriate knowledge. This study aimed to understand the level of HPV related knowledge among medical students and correlates in Southwest China in order to address any potential gap in their knowledge base. Methods: We conducted a cross-sectional survey among medical students in six selected universities across Southwest China based on a pretested questionnaire regarding HPV infection, cervical cancer, and HPV vaccines. Results: We successfully surveyed 1, 878 medical students, of whom 32.1% were males and 67.9% were females. Their mean age was 20.8 (standard deviation: 1.3) years. 91.8% of them were ethnic Han Chinese, and 43.8% were students in clinical medicine. While 76.5% had heard of HPV only 48.8% knew that the prevention of cervical cancer was possible through HPV vaccination. Only 42.9% of the male and 49.2% of the female students correctly answered over 10 out of 22 questions on HPV related knowledge. Female students appeared to know more about HPV and HPV vaccination (OR: 1.39; 95% CI: 1.11-1.75). In addition, the student knowledge improved with the grade (p<0.001). University courses were the most selected source of knowledge about HPV vaccination (66.4%). 83.6% of males and 91.4% of females were willing to have more HPV related education by experts (p<0.001). Only 10.1% of the students acknowledged that people had asked for their advice regarding HPV vaccination. Conclusions: Our survey indicates that medical students from Southwest China have poor knowledge of HPV and HPV vaccination, but are willing to receive more relevant information. Targeted education should be incorporated into school courses to inform these future health care providers and ensure success of programs for cervical cancer control and prevention.

• Journal of the Korean Society for Railway
• /
• v.19 no.5
• /
• pp.617-628
• /
• 2016

This paper proposes the structure of a smart transformer to improve the performance of the 60Hz main power transformer for rolling stock. The proposed smart transformer is a kind of solid state transformer that consists of semiconductor switching devices and high frequency transformers. This smart transformer would have smaller size than the conventional 60Hz main transformer for rolling stock, making it possible to operate AC electrified track efficiently by power factor control. The proposed structure employs a cascade H-Bridge converter to interface with the high voltage AC single phase grid as the rectifier part. Each H-Bridge converter in the rectifier part is connected by a Dual-Active-Bridge (DAB) converter to generate an isolated low voltage DC output source of the system. Because the AC voltage in the train system is a kind of medium voltage, the number of the modules would be several tens. To control the entire smart transformer, the inner DC voltage of the modules, the AC input current, and the output DC voltage must be controlled instantaneously. In this paper, a control algorithm to operate the proposed structure is suggested and confirmed through computer simulation.

• Proceedings of the KOSOMBE Conference
• /
• v.1996 no.11
• /
• pp.129-134
• /
• 1996

1차 년도 G-7 개발 과제로 수행된 자기 공명 진단 장치 (Magnetic Resonance Imaging System)의 개발 내용을 간략히 소개하였다. 성공적인 IT Compact 자기 공명 진단 장치의 완성을 위해 일차적으로 (1)RF (고주파), Gradient(경사 자계), Spectrometer 등의 Hard-ware 관련 MRI 핵심부분, (2) RF, Gradient, Spectrometer, Magnet 등의 각 Sub-system을 연결, 조합, 조정하여 하나의 체계적인 시스템으로 통합하고 운영하는 과정(System Integration), (3)사용자와 시스템을 연결하는 User Interface, Data Base Management, Real time 운영 SW 등과 (4)임상에 적용하여 구체적인 성능과 효용성을 확인하는 기술 등에 대하여 집중 연구하였다. 개발 방법은 (1)지난 16년간 국내에 축적 된 연구 개발 인력들을 최대한 활용하고 (2)연구 개발을 국제화 시켜 필요한 경우 부분별로 개발 인력을 해외에서 보완하고 (3)소수 정예 전문 인력 주의와 요소 기술 또는 중요 부품을 경쟁성 검토 후 필요 시 Out-sourcing 활용으로 최저의 비용으로 개발 기간을 최소화 하는 데 두었다. 개발된 1.0Tesla자기 공명 영상 장치는 미국 물리 학회에서 규격화한 Phantom및 임상 적용을 통하여 서울대 의대 연구 팀과 지속적으로 성능을 평가해 왔다. 개발된 시스템의 해상도는 $256{\times}256$ head 영상에서 1mm 이 하의 해상도를 가짐을 resolution phantom 을 통하여 확인할 수 있었고, $512{\times}512$ 영상에서 는 약 0.5 mm 의 물체를 분리 해냄으로써 외제 시스템들 보다 우수하게 평가 되었다. 차폐 경사코일의 Eddy current영향은2%이내로 촬영 시 영향은 거의 무시할 수 있었다. 또한, 개발된 영상 기법들, 즉 Multislice/Multi Echo, Oblique angle imaging, 64 Echo train을 갖는 고속 촬영 기술들이 자기 공명 장치에 장착되어 임상 적용에 문제가 없도록 하였다. 또한 20mT/m/Amp의 강력한 능동 차폐 경사 자계 코일(Active Shield Gradient Coil)을 기본 사양으로 하고, 수신단을 최대 6개로 확장토록 하여 2차년도의 초고속 촬영 기법(EPI) 및 Phased Array 코일 촬영이 가능토록 하였다. 1차 년도 개발 과제 수행 결과와 향후 개발 과제를 바탕으로 최종 목표인 국제 경쟁력이 있는 자기 공명 진단 장치 즉 기능과 영상의 질은 선진국 제품과 동일하거나 우수하되, 저가격을 구현한 상용화 제품이 완성되어, 첨단 의료기기로서 산업 구조 고도화에 기여하고 수입대체 뿐만 아니 라 수출을 통한 국익 창출과 국가의 기술을 통한 위상 제고에 기여되길 기대한다.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.23 no.6
• /
• pp.628-638
• /
• 2017

In response to difficult employment conditions in the maritime industry and a desire to expand their career options, domestic mates are persuing DP operator training at institutions both domestically and abroad based on their shipboard experience. However, since the offshore plant service industry has not yet been established in Korea, those seeking to enter this field have difficulty acquiring qualifications and most seek work overseas for offshore shipping companies. Individuals wishing to work as DP operators are likely to face more conservative recruitment processes with overseas offshore shipping companies, focusing on career language restrictions as they will be non-native speakers relative to the foreign company, difficulty living in a multi-cultural environment, and lack of systematic information on essential job requirements. For these reasons, domestic mates have difficulty seeking jobs. Therefore, this study analyzes the capabilities and qualification required to be a DP operator to provide basic data for developing NCS standards representing a minimum level of competency. These standards can be applied by the government to develop plans for professional training for DP operators. In study, job classifications, competency standards and career development paths for DP operators have been proposed along with joint use of DP training vessels, to train specialized DP instructors. An NCS export model led by the government to activate professional training for DP operators is also presented.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_2
• /
• pp.1709-1722
• /
• 2022

The rise in temperature induced by global warming caused in El Nino and La Nina, and abnormally changed the temperature of seawater. Rainfall concentrates in some locations due to abnormal variations in seawater temperature, causing frequent abnormal floods. It is important to rapidly detect flooded regions to recover and prevent human and property damage caused by floods. This is possible with synthetic aperture radar. This study aims to generate a model that directly derives flood-damaged areas by using modified U-NET and TerraSAR-X images based on Multi Kernel to reduce the effect of speckle noise through various characteristic map extraction and using two images before and after flooding as input data. To that purpose, two synthetic aperture radar (SAR) images were preprocessed to generate the model's input data, which was then applied to the modified U-NET structure to train the flood detection deep learning model. Through this method, the flood area could be detected at a high level with an average F1 score value of 0.966. This result is expected to contribute to the rapid recovery of flood-stricken areas and the derivation of flood-prevention measures.

• The Journal of the Acoustical Society of Korea
• /
• v.41 no.3
• /
• pp.359-366
• /
• 2022

In this paper, we present the Korean menu-ordering Sentence Text-to-Speech (TTS) system using conformer-based FastSpeech2. Conformer is the convolution-augmented transformer, which was originally proposed in Speech Recognition. Combining two different structures, the Conformer extracts better local and global features. It comprises two half Feed Forward module at the front and the end, sandwiching the Multi-Head Self-Attention module and Convolution module. We introduce the Conformer in Korean TTS, as we know it works well in Korean Speech Recognition. For comparison between transformer-based TTS model and Conformer-based one, we train FastSpeech2 and Conformer-based FastSpeech2. We collected a phoneme-balanced data set and used this for training our models. This corpus comprises not only general conversation, but also menu-ordering conversation consisting mainly of loanwords. This data set is the solution to the current Korean TTS model's degradation in loanwords. As a result of generating a synthesized sound using ParallelWave Gan, the Conformer-based FastSpeech2 achieved superior performance of MOS 4.04. We confirm that the model performance improved when the same structure was changed from transformer to Conformer in the Korean TTS.