• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2021년도 추계학술발표대회
• /
• pp.681-684
• /
• 2021

In this paper, we study image clustering without labeling using machine learning techniques. We proposed an unsupervised machine learning technique to design an image clustering model that automatically categorizes images into groups. Our experiment focused on inception convolutional neural networks (inception V3) with k-mean methods to cluster images. For this, we collect the public datasets containing Food-K5, Flowers, Handwritten Digit, Cats-dogs, and our dataset Rice Germination, and the owner dataset Palm print. Our experiment can expand into three-part; First, format all the images to un-label and move to whole datasets. Second, load dataset into the inception V3 extraction image features and transferred to the k-mean cluster group hold on six classes. Lastly, evaluate modeling accuracy using the confusion matrix base on precision, recall, F1 to analyze. In this our methods, we can get the results as 1) Handwritten Digit (precision = 1.000, recall = 1.000, F1 = 1.00), 2) Food-K5 (precision = 0.975, recall = 0.945, F1 = 0.96), 3) Palm print (precision = 1.000, recall = 0.999, F1 = 1.00), 4) Cats-dogs (precision = 0.997, recall = 0.475, F1 = 0.64), 5) Flowers (precision = 0.610, recall = 0.982, F1 = 0.75), and our dataset 6) Rice Germination (precision = 0.997, recall = 0.943, F1 = 0.97). Our experiment showed that modeling could get an accuracy rate of 0.8908; the outcomes state that the proposed model is strongest enough to differentiate the different images and classify them into clusters.

• 산업경영시스템학회지
• /
• 제46권1호
• /
• pp.55-67
• /
• 2023

Lightweight face recognition models, as one of the most popular and long-standing topics in the field of computer vision, has achieved vigorous development and has been widely used in many real-world applications due to fewer number of parameters, lower floating-point operations, and smaller model size. However, few surveys reviewed lightweight models and reimplemented these lightweight models by using the same calculating resource and training dataset. In this survey article, we present a comprehensive review about the recent research advances on the end-to-end efficient lightweight face recognition models and reimplement several of the most popular models. To start with, we introduce the overview of face recognition with lightweight models. Then, based on the construction of models, we categorize the lightweight models into: (1) artificially designing lightweight FR models, (2) pruned models to face recognition, (3) efficient automatic neural network architecture design based on neural architecture searching, (4) Knowledge distillation and (5) low-rank decomposition. As an example, we also introduce the SqueezeFaceNet and EfficientFaceNet by pruning SqueezeNet and EfficientNet. Additionally, we reimplement and present a detailed performance comparison of different lightweight models on the nine different test benchmarks. At last, the challenges and future works are provided. There are three main contributions in our survey: firstly, the categorized lightweight models can be conveniently identified so that we can explore new lightweight models for face recognition; secondly, the comprehensive performance comparisons are carried out so that ones can choose models when a state-of-the-art end-to-end face recognition system is deployed on mobile devices; thirdly, the challenges and future trends are stated to inspire our future works.

• 전력전자학회:학술대회논문집
• /
• 전력전자학회 2014년도 전력전자학술대회 논문집
• /
• pp.41-42
• /
• 2014

본 논문에서는 종전의 3상 PWM 정류기의 문제점을 극복하기 위하여 3상 Z-Source PWM 정류기를 제안하였다. 제안된 정류기는 시스템의 간소화, 안정화, 비용상승을 고려하여 상태관측기가 적용된 센서리스 제어 기법을 적용하였다. 3상 Z-Source PWM 정류기는 교류측 전압센서 없이 DC-link단 전압센서와 입력단 전류 센서만을 이용하여 교류측 전압을 추종한다. 제안된 3상 Z-Source PWM 정류기는 PSIM 시뮬레이션을 통하여 검증 하였다.

• 한국전기전자재료학회논문지
• /
• 제34권4호
• /
• pp.221-228
• /
• 2021

With the advent of the IoT (internet of things) era, there has been discussion on how to efficiently use various information from daily life. In academic and industrial society, various smart devices such as smart watches, smart phones, and smart glasses have been developed and commercialized for narrowing the physical/psychological distance with user information. According to recent developments of smart devices, the contemporary people have desired to check their body information and treat disease by themselves. According to the needs of the time, biological researches by phototherapy/monitoring have been actively conducted. Among various light sources, microLEDs have been spotlighted due to their superior optoelectric properties and stability. In this paper, we would like to review the state-of-the research results on the next-generation biological therapy devices via microLEDs.

• 대한방사성의약품학회지
• /
• 제6권2호
• /
• pp.165-170
• /
• 2020

Histone Deacetylases (HDACs) are enzymes that have control gene expression regulation and cell state. In additions, inhibitions of HDACs are associated with growth arrest, differentiation, or apoptosis of tumor cell. Thus HDAC inhibition is one of the interesting biological targets. A variety of HDAC inhibitors has been developed by many scientists, and some of chemical structures related with HDAC inhibitors were modified to give radiolabeled HDAC inhibitors for positron emission tomography (PET) study. In this highlight review, the development of radiolabeled HDAC inhibitors for PET study are described.

• 드라이브 ㆍ 컨트롤
• /
• 제21권1호
• /
• pp.53-58
• /
• 2024

With the recent development of intelligence and automation technologies for construction machinery, the demand for safety and efficiency of road-cutting operations has continued to increase. In response to this, a double-blade road cutter has been developed that can automatically cut roads. However, a double-blade road cutter has a load difference between the two blades due to the ground and wear conditions of the cutting blades. The difference in load between the two blades distorts the direction of travel of the cutter. In this study, a vision sensor-based driving guide technology was developed to correct the driving path of road cutters. In addition, we developed a load-sensing technology that detects blade loads in real-time and controls driving speed in the event of overload.

• 자원리싸이클링
• /
• 제33권1호
• /
• pp.22-30
• /
• 2024

2차원 물질 MXene의 전구체로 사용되는 Ti₃AlC₂ MAX 상 합성을 위해, 출발물질로써 타이타늄 (Ti) 스크랩을 활용하는 것은 경제적인 접근이 될 수 있다. 본 연구는 Ti 스크랩을 활용하여 Ti₃AlC₂ MAX상의 상분율 향상을 위한 합성 조건의 최적화를 수행하였다. Ti 스크랩으로부터 수소화-탈수소화(hydrogenation-dehydrogenation, HDH) 공정에 의해 제조된 Ti 분말의 산소 함량은 고상탈산(Deoxidation in solid state, DOSS) 공정을 통하여 효과적으로 감소되었다. 최적 합성 조건은 25 ~ 32 ㎛의 DOSS-Ti, Al, graphite 분말을 3:1.1:2의 몰 비율로 혼합하여 합성되었다. 이 때의 Ti₃AlC₂, TiC 및 Al₃Ti의 상분율은 각각 97.25%, 0.93%, 1.82%로 나타났으며, 25 ~ 45 ㎛의 Ti₃AlC₂ MAX 분말의 산소 함량은 4,210 ppm으로 확인되었다.

• 치위생과학회지
• /
• 제13권4호
• /
• pp.369-377
• /
• 2013

본 연구는 일부 대학생의 흡연실태와 흡연실태에 따른 흡연신념을 파악하여 대학생의 건강증진에 도움이 될 수 있는 흡연 예방 및 금연 프로그램 개발과 활용에 필요한 자료를 제공하고자 2012년 5월 1일부터 6월 1일까지 전라북도에 위치하고 있는 10개 대학교 학생 중 1,696명을 대상으로 분석한 결과 다음과 같은 결론을 얻었다. 1. 흡연실태는 흡연경험이 있는 학생은 23.7%였고, 처음 흡연한 시기는 고등학교가 42.1%로 가장 높았고, 흡연기간은 4년 이상이 46.1%로 가장 높게 나타났다. 2. 일반적 특성에 따른 흡연신념은 한 달 용돈이 30만원~39만원과, 40만원 이상에서 1.91점으로 가장 높았고(p<0.001), 흡연과 구강건강 관련성은 '전혀 그렇지 않다'가 2.00점으로 가장 높게 나타났다(p<0.001). 3. 흡연실태에 따른 흡연신념은 처음 흡연한 시기에서 초등학교가 2.66점으로 가장 높았고, 친한 친구 5명 중 흡연하는 친구는 4명이 2.60점으로 가장 높게 나타났다(p<0.05). 4. 흡연의존도와 금연의도에 따른 흡연신념은 흡연의존도에서 '매우 그렇다'가 2.75점으로 가장 높았고(p<0.001), 금연의도는 '금연하지 않음'이 2.68점으로 가장 높게 나타났다(p<0.01).

• 한국전기전자재료학회논문지
• /
• 제34권6호
• /
• pp.454-460
• /
• 2021

In this work, we synthesized alloy-core InZnP quantum dots, which are more efficient than single-core InP quantum dots, using a solution process method. The effect of synthesis conditions of alloy core on optical properties was investigated. We also investigated the conditions that make up the gradient shell to minimize defects caused by lattice mismatch between the InZnP core and ZnS is 7.7%. The stable synthesis temperature of the InZnP alloy core was 200℃. Quantum dots consisting of three layered ZnSe gradient shell and single layered ZnS exhibited the best optical property. The properties of quantum dots synthesized in 100 ml and in 2,000 ml flasks were almost equal.

• 공학교육연구
• /
• 제23권6호
• /
• pp.33-39
• /
• 2020

This paper analyzes the current state of operation and the effect of introduction of the single accreditation system, which was fully introduced in 2016 by the Accreditation Board of Engineering Education Korea(ABEEK). Because Engineering Education Accreditation are program-based certification, and department operates accreditation programs and non-accreditation programs, if a student under the accreditation program fails to graduate from the accreditation program, he or she must complete the transfer to a non-accreditation program at least one year before graduation. Since 2016, when the single accreditation system was introduced, freshmen of department have allowed non-accreditation graduation only for reasons prescribed separately by the program's regulations or guidelines. In order to identify and reflect the operational status and effectiveness of the single accreditation system in the 2020 accreditation review, ABEEK conducted a complete survey on the current status of graduates in February 2020 among 157 programs at 26 universities receiving the 2020 accreditation review. The results of this study are expected to be used as basic data for the improvement of the single accreditation system in the future by examining the effects of the single accreditation system implemented after 2016 and considering the problems at the site due to the introduction of the single accreditation system of programs that ABEEK had not previously expected.