• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.9
• /
• pp.2189-2196
• /
• 2014

In this paper, a VLIW(Very Long Instruction Word) vision coprocessor which can efficiently accelerate computer vision algorithm for automotive is designed. The VLIW coprocessor executes four instructions per clock cycle via 8-stage pipelined structure and has 36 integer and floating-point instructions to accelerate computer vision algorithm for pedestrian detection. The processor has about 300-MHz operating frequency and about 210,900 gates under 45nm CMOS technology and its estimated performance is 1.2 GOPS(Giga Operations Per Second). The vision system composed of vision primitive engine and eight VLIW coprocessors can execute pedestrian detection at 25~29 frames per second(FPS). Because the VLIW coprocessor has high detection rate and loosely coupled interface with host processor, it can be efficiently applicable to a wide range of vision applications.

• International Journal of Contents
• /
• v.14 no.2
• /
• pp.24-29
• /
• 2018

In this paper, we propose an efficient traffic light detection (TLD) method as an advanced driver assistance system (ADAS) for elderly drivers. Since an increase in traffic accidents is associated with the aging population and an increase in elderly drivers causes a serious social problem, the provision of ADAS for older drivers via TLD is becoming a necessary(Ed: verify word choice: necessary?) public service. Therefore, we propose an economical TLD method that can be implemented with a simple black box (built in camera) and a smartphone in the near future. The system utilizes a color pre-processing method to differentiate between the stop and go signals. A mathematical morphology algorithm is used to further enhance the traffic light detection and a circular Hough transform is utilized to detect the traffic light correctly. From the simulation results of the computer vision and image processing based on a proposed algorithm on Matlab, we found that the proposed TLD method can detect the stop and go signals from the traffic lights not only in daytime, but also at night. In the future, it will be possible to reduce the traffic accident rate by recognizing the traffic signal and informing the elderly of how to drive by voice.

• Journal of the Korea Society of Computer and Information
• /
• v.23 no.11
• /
• pp.85-93
• /
• 2018

In recent years, there has been an increasing number of ways to distribute document-based malicious code using vulnerabilities in document files. Because document type malware is not an executable file itself, it is easy to bypass existing security programs, so research on a model to detect it is necessary. In this study, we extract main features from the document structure and the JavaScript contained in the stream object In addition, when JavaScript is inserted, keywords with high occurrence frequency in malicious code such as function name, reserved word and the readable string in the script are extracted. Then, we generate a machine learning model that can distinguish between normal and malicious. In order to make it difficult to bypass, we try to achieve good performance in a black box type algorithm. For an experiment, a large amount of documents compared to previous studies is analyzed. Experimental results show 98.9% detection rate from three different type algorithms. SVM, which is a black box type algorithm and makes obfuscation difficult, shows much higher performance than in previous studies.

• The Journal of Korea CHUNA Manual Medicine
• /
• v.5 no.1
• /
• pp.19-29
• /
• 2004

Objectives : Visual check and X-ray are commonly used by chiropractors to estimate ieg length inequality, This study have three categories: diagnosis for anatomic leg length inequality; difference between anatomic and functional leg length inequality; theraphies for anatomic or functional leg length inequality. Methods : We referred to a PubMed site by using word of 'leg length [JU] J Manipulative Physiol Ther', only items with abstracts. Results : We searched 26 articles in J Manipulative Physiol Ther with the key word-Ieg length. Conclusion : 1. Radiographs were most accurate and commonly used by chiropractors to measure anatomic leg length inequality, clinically wood block, tape measure, visual check are acceptable. 2. There was no article about difference between anatomic and functional leg length inequality. 3. Heel lift was commonly used with conservative theraphy for anatomic leg length Inequality. 4. Chiropractors have not yet proved that the supposed positive effects are a result of a reduction of subluxation, The detection of the manipulative lesion in the sacroiliac joint depends on valid and reliable tests, Because such tests have not been established, the presence of the manipulative lesion remains hypothetical. Great effort is needed to develop, establish and enforce valid and reliable test procedures.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.6 no.2
• /
• pp.330-338
• /
• 2002

A speech recognition chip that can recognize a small vocabulary as a word-level has been designed. It is composed of EPD(Start and End-point detection) block, LPC block, DTW block and external memory interface block. It is made of 126,938 gates on 4x4mm2 area with a CMOS 0.35um TLM process. The speed of the chip varies from 5MHz to 60MHz because of its specific hardware designed for the purpose. It can compare 100,000 voices as a small vocabulary which has approximately 50∼60 frames at the clock of 5MHz and also up to 1,200,000 voices at the clock of 60MHz.

• Korean Journal of Biological Psychiatry
• /
• v.20 no.3
• /
• pp.104-110
• /
• 2013

Objectives : The aim of this study was to examine the prospective impact of the apolipoprotein E (APOE) ${\varepsilon}4$ on cognitive performance in the community-dwelling elderly individuals with Alzheimer's disease (AD). Methods : The total number of subjects was 30 (12 men and 18 women) who were diagnosed with AD from a Korean project of "Early Detection of Dementia". People aged 65-85 years were included in the analysis. The eight neuropsychological domains from the Korean version of Consortium to Establish a Registry of Alzheimer's Disease (CERAD-K) were conducted to test subjects. They have been followed at 24-month intervals with the same assessments at each interval. Their cognitive performance at 2 year intervals was compared by the occurrence of the APOE ${\varepsilon}4$. Results : The impact of ${\varepsilon}4$ allele was significant in the Word List Memory Test (WLMT, F = 4.345, df = 1, p = 0.021) and Word List Recall Test (WLRT, F = 5.569, df = 1, p = 0.033). Conclusions : The APOE ${\varepsilon}4$ allele was significantly correlated especially with verbal episodic memory domain in community-dwelling elders diagnosed with AD.

• Food Science and Industry
• /
• v.54 no.3
• /
• pp.160-170
• /
• 2021

A recent advance in communication technologies accelerates the spread of food safety issues once presented by the news media. To respond to those safety issues and take steps in a timely manner, automatically detecting related information from the news data matters. This work presents an AI-based system that detects risk information within a food-related news article. Experts in food safety areas participated in labeling risk information from the food-related news articles; we acquired 43,527 articles in which food names and risk information are marked as labels. Based on the news document, our system automatically detects food names and risk information by analyzing similarities between words within a text by leveraging learned word embedding vectors. Our AI-based system shows higher detection accuracy scores over a non-AI rule-based system: achieving an absolute gain of +32.94% in F1 for the food name category and +41.53% for the risk information category.

• Annual Conference of KIPS
• /
• 2009.04a
• /
• pp.1277-1280
• /
• 2009

본 논문은 ITU-T 권고안 J-38 부록 B에 명시된 전송방식의 분석 및 시뮬레이션을 토대로 성능을 분석 하였으며 FPGA 구현시 야기되는 문제점을 나타내고, 해결방안을 제시하였다. 구현상의 문제점으로는 크게 두 가지로 분류되는데, 첫째로 다양한 부호화 방식과 변조방식 그리고 심볼 단위 및 비트 단위의 처리로 인해 많은 클럭수를 요구하는데 본 논문에서는 읽기/쓰기 메모리를 이용하여 필요한 클럭수를 줄였다. 둘째로는 펑쳐링 부호화된 TCM 복호기에 펑처링 패턴에 정확한 동기를 얻지 못하면 프레임 동기 심볼인 UW(Unique sync-Word)를 획득하지 못하여 모든 데이터가 에러 처리되기 때문에 본 논문에서는 펑처링 패턴과 UW 심볼의 동기를 맞추는 알고리즘을 제시하였다. 이러한 알고리즘 분석 및 구현상의 문제점 해결을 토대로 본 논문에서는 ITU-T J38 annex B의 하향 스트림 채널 부호화 시스템을 VHDL 언어를 사용하여 FPGA 칩에 직접 구현하였다.

• Korean Journal of Cognitive Science
• /
• v.18 no.1
• /
• pp.15-34
• /
• 2007

Two experiments were conducted to examine whether the word superiority effect (WSE) can be affected by visual familiarity of Hangout font and visual identicalness between target letter and test letter. In experiment 1, Using Batang-font as a target font, it was tested whether the consistency between test font and target font could affect the WSE across different display time conditions. In experiment 2, using the Batang-font and Sebeul-font, the effort of visual familiarity on WSE were investigated. The results revealed that the WSE was found regardless the consistency between target and test letter. In addition, it was found that visual familiarity could affect overall detection rate, but not WSE. These results suggested that the WSE on Hanguel is independent of visual properties.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.16 no.10
• /
• pp.7211-7218
• /
• 2015

Tens of thousands of malicious codes are generated on average in a day. New types of malicious codes are surging each year. Diverse methods are used to detect such codes including those based on signature, API flow, strings, etc. But most of them are limited in detecting new malicious codes due to bypass techniques. Therefore, a lot of researches have been performed for more efficient detection of malicious codes. Of them, visualization technique is one of the most actively researched areas these days. Since the method enables more intuitive recognition of malicious codes, it is useful in detecting and examining a large number of malicious codes efficiently. In this paper, we analyze the relationships between malicious codes and Native API functions. Also, by applying the word cloud with text mining technique, major Native APIs of malicious codes are visualized to assess their maliciousness. The proposed malicious code analysis method would be helpful in intuitively probing behaviors of malware.