• 한국멀티미디어학회논문지
• /
• 제25권8호
• /
• pp.1049-1058
• /
• 2022

HAR(Human Action Recognition) such as anomaly and object detection has become a trend in research field(s) that focus on utilizing Artificial Intelligence (AI) methods to analyze patterns of human action in crime-ridden area(s), media services, and industrial facilities. Especially, in real-time system(s) using video streaming data, HAR has become a more important AI-based research field in application development and many different research fields using HAR have currently been developed and improved. In this paper, we propose and analyze a deep-learning-based HAR that provides more efficient scheme(s) using an intelligent AI models, such system can be applied to media services using RGB video streaming data usage without feature extraction pre-processing. For the method, we adopt Slowfast based on the Deep Neural Network(DNN) model under an open dataset(HMDB-51 or UCF101) for improvement in prediction accuracy.

• 스마트미디어저널
• /
• 제12권1호
• /
• pp.54-61
• /
• 2023

Pricing, as one of the most important aspects of a business, should be taken seriously. Whatever affects a company's pricing system tends to affect its profits and losses as well. Currently, many manufacturing companies fix product prices manually by members of an organization's management team. However, due to the imperfect nature of humans, an extremely low or high price may be fixed, which is detrimental to the company in either case. This paper proposes the development of a fuzzy-based price expert system (Expert Fuzzy Price (EFP)) for manufacturing companies. This system will be able to recommend appropriate prices for products in manufacturing companies based on four major pricing strategic goals, namely: Product Demand, Price Skimming, Competition Price, and Target population.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2023년도 춘계학술발표대회
• /
• pp.665-666
• /
• 2023

딥러닝은 인간이 탐지하기 어려운 데이터의 특징 및 패턴을 인지하고, 이들을 학습하여 데이터를 분류 및 예측할 수 있는 기술이다. 그러나 딥러닝 모델을 잘 학습시키기 위해서는 고품질의 대용량 데이터와 이들을 처리할 수 있는 방대한 컴퓨터 자원이 요구되는 것이 일반적이다. 따라서 소량의 데이터만이 존재하는 분야나 컴퓨터 자원이 한정되어 있는 상황에서는 딥러닝을 적용하기 어렵다. 본 논문에서는, 소량의 데이터로도 모델을 자신들의 태스크에 맞게 최적화시킬 수 있는 메타러닝에 대해 소개하고, 메타 러닝 기법들의 방향에 따른 Metric-Based, Model-Based 및 Optimization 기반 모델들에 대해 소개하고, 앞으로 나아가야 할 연구 방향에 대해 제시한다.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2023년도 제67차 동계학술대회논문집 31권1호
• /
• pp.323-324
• /
• 2023

본 논문에서는 다양한 복원 인공지능 알고리즘 중 하나인 3차원 복원 기술은 실제로 존재하는 물체의 2차원적인 픽셀을 3차원의 형태로 구현하여 형상화한다. 정확한 3차원 정보 처리가 요구됨에 따라 포인트 클라우드로 표현되는 데이터를 통해 정확한 쿨체의 크기 정보나 좌표 정보를 표시할 수 있다. 데이터의 픽셀을 분석하여 3차원의 형태로 구현할 것을 정의하는 복셀화(Voxelization) 알고리즘 전처리 과정을 통해 3차원 복원 기술 3D-GAN 활용으로 3차원 형태 형상화를 하였다. 본 논문에서는 3차원 복원 알고리즘 통하여 2차원 포인트 클라우드를 분석해 3차원 형태로 복원하는 기술에 대한 설명한다.

• 스마트미디어저널
• /
• 제13권2호
• /
• pp.85-94
• /
• 2024

Artificial Intelligence of Things (AIoT), which combines AI and the Internet of Things (IoT), has recently gained popularity. Deep neural networks (DNNs) have achieved great success in many applications. Deploying complex AI models on embedded boards, nevertheless, may be challenging due to computational limitations or intelligent model complexity. This paper focuses on an AIoT-based system for smart sewing automation using edge devices. Our technique included developing a detection model and a decision tree for a sufficient testing scenario. YOLOv5 set the stage for our defective sewing stitches detection model, to detect anomalies and classify the sewing patterns. According to the experimental testing, the proposed approach achieved a perfect score with accuracy and F1score of 1.0, False Positive Rate (FPR), False Negative Rate (FNR) of 0, and a speed of 0.07 seconds with file size 2.43MB.

• 한국생물공학회:학술대회논문집
• /
• 한국생물공학회 2000년도 추계학술발표대회 및 bio-venture fair
• /
• pp.273-276
• /
• 2000

Marine fungus, Thraustochytrium aureum ATCC 34034, was incubated in artificial sea water media in order to produce docosahexaenoic acid(DHA). Cultures were performed at $24^{\circ}C$ in light for 13 days with orbital shaker at 100rpm. Maximum cell productivity of 1.34g/L and DHA yield of 41.4 mg/L were obtained by using this method, which is almost the twice level of DHA yield obtained for the strains reported previously. This strain did produce much more DHA after sporangium disruption.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2002년도 ITC-CSCC -3
• /
• pp.2035-2038
• /
• 2002

The necessity of the filter coefficients compensation for the LTJ adaptive filter was explained generally and easily by analyzing it with respect to the time-varying transform domain adaptive filter. And also the reduction method of computational complexity for filter coefficients compensation was proposed and its effectiveness was verified through experiments using artificial and real speech signals. The proposed adaptive filter reduces the computational complexity for filter coefficients compensation by 95%, and when the filter is applied to the acoustic echo canceller with 1000 taps, the total complexity is reduced by 82%

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송∙미디어공학회 2022년도 하계학술대회
• /
• pp.1218-1221
• /
• 2022

최근 전세계의 탄소중립 요구에 따른 에너지 효율 증대를 통한 에너지 절감을 위한 효율성 관련 연구가 확대되고 있다. 방송과 미디어 분야에는 에너지 효율이 더욱 시급하다. 이에 본 연구에서는 효율적인 에너지 시스템 구축을 위해 난방 에너지 시계열 데이터를 기반으로 한 수요 예측 모델을 선정하고, 설명하는 인공지능 모델을 도입하여 수요 예측에 영향을 미치는 원인을 파악하는 프레임워크를 제안한다.

• 부모자녀건강학회지
• /
• 제6권1호
• /
• pp.70-81
• /
• 2003

The otitis media has the highest rate of occurrence in 6-36 month after birth, and relatively high rate of occurrence to six years old. It is one of major cause for hearing loss problem because of its high rate of recurrence. The researcher observed at bed that feeding posture and feeding habit at night have some relationship with occurrence of otitis media. I found that medical research in this area was very weak, and it caused her to concentrate her research on the topic. This paper was undertook a retrospective case-control study to find out the relationship between the mode of feeding and the occurrence of otitis media among the otitis media patients and the healthy persons. As the controling groups, the researcher chose fifty eight mothers whose children received the treatment of otitis media(6-36 months after birth) from the pediatrics clinic of two general hospitals in Seoul, as the patient group, and forty five mothers who consulted with their child to well baby clinic at the general hospital and kindergarten in Seoul as the control group. The results of the survey can be summarized as followed : 1. "Otitis media patient group may have more lying posture mothers while feeding than the healthy control group". The first hypothesis is supported by the fact that the children who have lying posture feeding habit mothers are more easily exposed to otitis media than those who have sitting posture feeding habit mothers as the analysis show : $X^2=8.142$, p = .017 2. "Otitis media patient group may have higher rate of sleeping habit with milk bottle on the month at night than the healthy control group does". The second hypothesis was supported by the fact that the patients group has higher rate of expected sleeping habit than the healthy group dose as the analysis show : $X^2=4.35$, p = .037 3. "Otitis media can be found more in the artificial feeding group than in the healthy children group". The hypothesis is rejected by the analysis, $X^2=1.550$, p = .416. Though the hypothesis is not supported by this research, we need to encourage mothers to feed mother's breast milk on the ground that healthy group has mother's milk feeding tendency, and the best food for infants is their mother's milk, and psychological-sociological effects of breast feeding os quite good for children. It will improve the health of infants. It can be summarized as follows : Otitis media is more concerned with posture of feeding and night feeding habits than feeding modes. On the basis of this survey, feeding education problem for mothers of infants need to be developed. In the education, sitting-embracing posture of feeding, mother's breast feeding, and prohibition of children's lying posture feeding at night need to be emphasized.

• 한국환경복원기술학회지
• /
• 제7권4호
• /
• pp.98-109
• /
• 2004

This study was conducted to develop revegetation methods for the restoration of the steep slopes by recycling of bark compost and mushroom media. In general, bark compost and mushroom media can be used as soil media for the restoration works, because they can increase infiltration of rainfall and give enough porous to breathe and elongate for the root growth as well as environmental value. This experiment was carried out to know the effect of soil media composed by different ratio of mushroom media for the restoration of steep rock-exposed slopes, and to certificate how soil media(PEC) will be effective to germinate and grow for native plants. The main results of the study are summarized as follows; 1. In percent coverage, the soil media PEC1 is more valuable than PEC2. The seed mixtures recommended by Government of Transportation and Construction can be used at PEC1 and PEC2, but it will be more useful if the total amount of seed are reduced and seed mixtures are altered in a direction of native plants. 2. The soil media is under 20 mm tested by Yamanaka Hardiness Tester which is available for the seed germination and growth. 3. The surface cracks are not occurred in PEC1 and PEC2, but more than 30 cracks per 1 square meter are occurred at soil media which is constructed by normal soil-seed-fertilizer hydro-seeding methods. 4. The soil moisture contents are over 20 percent level during 15 day. Such moisture content in soil media will be effective for the plant growth. By using Terra-Control, PEC can maintain enough soil moisture. 5. The eroded soils from $60^{\circ}$ slopes by artificial rainfall with the intensity of 20 mm/hr in one day after seeding are estimated under 1%. By the results of erosion test, it comes to the conclusion that soil media of PEC can be adapted at steep rock exposed slopes.