• Imaging Science in Dentistry
• /
• 제52권3호
• /
• pp.239-244
• /
• 2022

Purpose: Despite the proliferation of numerous morphometric and anthropometric methods for sex identification based on linear, angular, and regional measurements of various parts of the body, these methods are subject to error due to the observer's knowledge and expertise. This study aimed to explore the possibility of automated sex determination using convolutional neural networks(CNNs) based on lateral cephalometric radiographs. Materials and Methods: Lateral cephalometric radiographs of 1,476 Iranian subjects (794 women and 682 men) from 18 to 49 years of age were included. Lateral cephalometric radiographs were considered as a network input and output layer including 2 classes(male and female). Eighty percent of the data was used as a training set and the rest as a test set. Hyperparameter tuning of each network was done after preprocessing and data augmentation steps. The predictive performance of different architectures (DenseNet, ResNet, and VGG) was evaluated based on their accuracy in test sets. Results: The CNN based on the DenseNet121 architecture, with an overall accuracy of 90%, had the best predictive power in sex determination. The prediction accuracy of this model was almost equal for men and women. Furthermore, with all architectures, the use of transfer learning improved predictive performance. Conclusion: The results confirmed that a CNN could predict a person's sex with high accuracy. This prediction was independent of human bias because feature extraction was done automatically. However, for more accurate sex determination on a wider scale, further studies with larger sample sizes are desirable.

• 산업경영시스템학회지
• /
• 제45권3호
• /
• pp.49-56
• /
• 2022

ROK Navy intends to secure the Korean-type aircraft carrier in order to effectively prepare for various future security threats. In general, the Korean national competency is considered to be at the level of having an aircraft carrier, but it is unclear what scale aircraft carrier would be appropriate. In this study, the efficiency was evaluated through the relative comparison between national competency(national power, economic power) and the scale of aircraft carriers, and the optimal scale of the Korean-type aircraft carrier that could be acquired was presented. A DEA(Data Envelopment Analysis) model was applied to aircraft carriers(19 aircraft carriers in 11 countries) currently in operation and scheduled to be possessed in the world. As input variables, CINC(Composite Index of National Capability) and GDP(Gross Domestic Product), which are the most widely used as indicators of national and economic power, and as output variables, the full-load displacement, length, and width of aircraft carriers were selected. ARIMA(short-term within 5 years) and simple regression(long-term over 5 years) were used to estimate the future national competency of each country at the time of aircraft carriers acquisition. The relative efficiency score of the Korean-type aircraft carrier currently being evaluated is 1.062, and it was evaluated as small-scale aircraft carrier compared to the national competency. Based on Korean national competency, the optimal scale of the Korean-type aircraft carrier calculated by aggregating benchmark groups, is 58,308.1 tons of full-load displacement, 279.4m in length, and 68.3m in width.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2023년도 학술발표회
• /
• pp.184-184
• /
• 2023

Globally, agriculture is one of the largest consumers and polluters of water resources, contributing to the unsustainable use of limited water resources. To reduce the resource use and environmental footprints associated with current and future food systems, researchers and policy makers have recommended the transition to sustainable and healthier diets and the reduction of food loss and waste along the food supply chain. However, there is limited information on the synergistic effects and trade-offs of adopting the two measures. In this study, we assessed the water-saving potential of the two measures in South Korea using environmentally extended input-output relying on the EXIOBASE database for the reference year 2020, along with scenario analysis to model the potential outcomes. Specifically, we analyzed scenarios where meat consumption was reduced by 30% and 50% and in combination with a 50% reduction in food waste at the consumption stage for each scenario. According to our findings, by considering individual measures of dietary change and food waste reduction, shifting to a diet with 30% and 50% less meat consumption could lead to reduction in water footprint by 6.9% and 7.5%, respectively, while 50% reduction in food waste at the consumption stage could save about 14% of water footprint. However, the synergistic effects of the two measures such as 30% less meat consumption and 50% food waste reduction, and 50% less meat consumption and 50% food waste reduction result to 20% and 24% reductions in water footprint, respectively. Moreover, our findings also showed that increasing food consumption with high environmental impacts could promote resources use inefficiency when waste occurs. Thus, policy strategies that address synergistic effects of both dietary change and food waste reduction should be strengthened to achieve sustainable food system. International and national policies can increase resource efficiency by utilizing all available reduction potentials while considering strategies interactions.

• Geomechanics and Engineering
• /
• 제33권3호
• /
• pp.279-289
• /
• 2023

Brittleness index (BI) is an important property of rocks because it is a good index to predict rockburst. Due to its importance, several empirical and soft computing (SC) models have been proposed in the literature based on the punch penetration test (PPT) results. These models are very important as there is no clear-cut experimental means for measuring BI asides the PPT which is very costly and time consuming to perform. This study used a novel Multivariate Adaptive regression spline (MARS), M5P, and white-box ANN to predict the BI of rocks using the available data in the literature for an improved BI prediction. The rock density, uniaxial compressive strength (σ_c) and tensile strength (σ_t) were used as the input parameters into the models while the BI was the targeted output. The models were implemented in the MATLAB software. The results of the proposed models were compared with those from existing multilinear regression, linear and nonlinear particle swarm optimization (PSO) and genetic algorithm (GA) based models using similar datasets. The coefficient of determination (R²), adjusted R² (Adj R²), root-mean squared error (RMSE) and mean absolute percentage error (MAPE) were the indices used for the comparison. The outcomes of the comparison revealed that the proposed ANN and MARS models performed better than the other models with R² and Adj R² values above 0.9 and least error values while the M5P gave similar performance to those of the existing models. Weight partitioning method was also used to examine the percentage contribution of model predictors to the predicted BI and tensile strength was found to have the highest influence on the predicted BI.

• 대한조선학회논문집
• /
• 제60권3호
• /
• pp.165-174
• /
• 2023

Unmanned Surface Vehicle (USV)'s berthing and unberthing is the most difficult maneuvering tasks and have the highest risk of accidents. In this paper, we designed a berthing/unberthing control algorithm given human joystick command for an USV equipped with a waterjet and a bow thruster. The berthing and unberthing maneuvers are performed remotely by a joystick operator at the Ground Control Center (GCC) where the status of USV and environmental situation can be monitored. We interpret the human joystick commands into USV's desired speed, yaw rate, and heading angle commands. next, we developed a control algorithm for the desired target values of MIMO actuators (engine speed, bucket step, nozzle angle, and bow thruster state) to follow the interpreted commands. The validity of the control algorithm is confirmed through simulations and sea trials at Gwang Am port.

• 한국컴퓨터정보학회논문지
• /
• 제28권10호
• /
• pp.37-53
• /
• 2023

인공지능에 기반한 지능형 지휘통체체계는 복잡하고 방대한 전장정보와 전술 데이터들을 학습모델을 통해 자동으로 융합 및 추출하여 전장상황을 분석한다. 지휘관은 지능형 지휘통제체계의 상황분석 결과를 제공받아 전장인식이 가능하여 의사결정을 지원할 수 있다. 의사결정지원에 특화된 결과를 지휘관에게 제공하기 위해서는 인공지능을 학습하기 위한 실 전장상황과 유사한 전장상황분석 데이터셋 생성이 필요하다. 본 논문은 기존 선행연구인 '인공지능 기반 전장상황분석을 위한 가상 전장상황 데이터 셋 생성 연구'의 다음 단계의 데이터셋 구축 방법 연구로 지휘관의 의사결정지원 및 미래 전장인식을 위해 최종적인 전장상황분석 결과에 필요한 데이터셋을 생성하는 방안에 대해 제안하였다. 전장상황 분석용 학습 데이터셋 생성도구 SW를 설계 및 구현하였고, 구현한 SW를 이용하여 데이터 레이블 작업을 진행하였다. Siamese Network 학습모델을 이용하여 구축한 데이터셋을 입력하고, 후처리 알고리즘을 활용한 출력 결과를 도출하여 생성한 데이터셋을 검증하였다.

• Archives of Plastic Surgery
• /
• 제51권1호
• /
• pp.30-35
• /
• 2024

Background Breast aesthetics evaluation often relies on subjective assessments, leading to the need for objective, automated tools. We developed the Seoul Breast Esthetic Scoring Tool (S-BEST), a photometric analysis software that utilizes a DenseNet-264 deep learning model to automatically evaluate breast landmarks and asymmetry indices. Methods S-BEST was trained on a dataset of frontal breast photographs annotated with 30 specific landmarks, divided into an 80-20 training-validation split. The software requires the distances of sternal notch to nipple or nipple-to-nipple as input and performs image preprocessing steps, including ratio correction and 8-bit normalization. Breast asymmetry indices and centimeter-based measurements are provided as the output. The accuracy of S-BEST was validated using a paired t-test and Bland-Altman plots, comparing its measurements to those obtained from physical examinations of 100 females diagnosed with breast cancer. Results S-BEST demonstrated high accuracy in automatic landmark localization, with most distances showing no statistically significant difference compared with physical measurements. However, the nipple to inframammary fold distance showed a significant bias, with a coefficient of determination ranging from 0.3787 to 0.4234 for the left and right sides, respectively. Conclusion S-BEST provides a fast, reliable, and automated approach for breast aesthetic evaluation based on 2D frontal photographs. While limited by its inability to capture volumetric attributes or multiple viewpoints, it serves as an accessible tool for both clinical and research applications.

• 한국임상수의학회지
• /
• 제23권2호
• /
• pp.105-110
• /
• 2006

2003년 5월부터 2005년 11월까지 산란계의 전염성기관지염(IB) 예찰 프로그램에 등록한 농장에 대한 역학조사에서 얻은 자료에 근거하여 IB 감염을 확인할 수 있는 모형을 구축하기 위하여 16개의 입력 뉴런, 3 개의 은닉 뉴런, 1개의 출력 뉴런으로 구성된 3층 인공신경망 모형을 개발하였다. 총 86개의 계군 중 77개는 훈련자료에 할당하고 나머지 9개는 검정자료로 무작위로 할당하여 back-propagation algorithm으로 신경망 훈련을 수행하였다. 입력 뉴런은 산란계군의 특성, 사양관리, 계군의 크기 등 16개의 역학조사 항목을 사용하였으며 출력 뉴런은 IB 감염의 유무로 투입하였다. 훈련된 신경망을 검정자료에 적용하여 민감도와 특이도를 산출하였으며 진단의 정확도는 receiver operating characteristic (ROC) 곡선을 사용하여 곡선 밑의 면적(AUC)을 계산하여 평가하였다. 입력 뉴런의 특성과 훈련모수를 변경하면서 다양한 신경망을 구성하였으며 최적의 신경망으로 확인된 IBV_D1 신경망의 경우 훈련자료에 대하여 77건 중 73건을 올바르게 판단하여 94.8%의 정확도를 보였다. 민감도와 특이도는 각각 95.5% (42/44, 95% CI, 84.5-99.4)와 93.9% (31/33, 95% CI, 79.8-99.3)로 나타났다. 훈련된 신경망을 검정자료에 적용하여 ROC 곡선을 작성한 결과 AUC는 전체의 94.8% (SE=0.086, 95% CI 0.592-0.961)를 차지하는 우수한 모형으로 나타났다. ROC 곡선에서 기준을 0.7149 이상으로 판단할 때 진단의 정확도가 88.9%로 가장 높았으며 100%의 민감도를 달성하였다. 이러한 민감도와 특이도에서 44%의 IB 유병률을 가정할 때 IBV_D1 모형은 80%의 양성예측도와 100%의 음성예측도를 보였다. 이러한 소견에 근거할 때 본 연구에서 구축한 신경망 모형은 산란계군에서 IB의 존재를 확인하기 위한 목적에 성공적으로 응용될 수 있을 것으로 판단되었다.

• 제어로봇시스템학회논문지
• /
• 제16권12호
• /
• pp.1150-1158
• /
• 2010

In this study, the Polynomial-based Radial Basis Function Neural Networks is proposed as one of the recognition part of overall face recognition system that consists of two parts such as the preprocessing part and recognition part. The design methodology and procedure of the proposed pRBFNNs are presented to obtain the solution to high-dimensional pattern recognition problem. First, in preprocessing part, we use a CCD camera to obtain a picture frame in real-time. By using histogram equalization method, we can partially enhance the distorted image influenced by natural as well as artificial illumination. We use an AdaBoost algorithm proposed by Viola and Jones, which is exploited for the detection of facial image area between face and non-facial image area. As the feature extraction algorithm, PCA method is used. In this study, the PCA method, which is a feature extraction algorithm, is used to carry out the dimension reduction of facial image area formed by high-dimensional information. Secondly, we use pRBFNNs to identify the ID by recognizing unique pattern of each person. The proposed pRBFNNs architecture consists of three functional modules such as the condition part, the conclusion part, and the inference part as fuzzy rules formed in 'If-then' format. In the condition part of fuzzy rules, input space is partitioned with Fuzzy C-Means clustering. In the conclusion part of rules, the connection weight of pRBFNNs is represented as three kinds of polynomials such as constant, linear, and quadratic. Coefficients of connection weight identified with back-propagation using gradient descent method. The output of pRBFNNs model is obtained by fuzzy inference method in the inference part of fuzzy rules. The essential design parameters (including learning rate, momentum coefficient and fuzzification coefficient) of the networks are optimized by means of the Particle Swarm Optimization. The proposed pRBFNNs are applied to real-time face recognition system and then demonstrated from the viewpoint of output performance and recognition rate.

• 한국통신학회논문지
• /
• 제11권3호
• /
• pp.179-196
• /
• 1986

본 논문에서는 통계적 패킷 음성/데이터 다중화기의 성능을 연구하였다. 성능해석은 음성과 데이터가 서로 분리된 한정된 queue를 사용하고, 전송에 있어서 음성이 데이터보다 우선권을 갖는 것을 가정하고, 다중화기의 출력 link를 시간 slot단위로 나누고 음성은 (M+1)-state의 Markov Process로, 데이터는 Poisson process로 modeling 하여 수행하였다. 전송시 음성신호가 데이터 신호보다 우선권을 가지므로 음성의 queueing behavior는 data에 거의 영향을 받지 않는다. 다라서 본 연구에서는 음성의 queueing behavior를 먼저 해석한 다음 data의 queueing behavior를 해석하였다. 패킷 음성 다중화기의 성능 해석은 입력상태와 buffer의 점유를 2차원의 Markov chain을 가지고 formulation하였고, 집적된 음성/data의 다중화기는 data를 추가한 3차원 Markov chain으로 하였다. 이러한 model을 사용하여 Gauss-Seidel방법으로 결과를 얻고 simulation으로 입증하였다. 이들 결과로 부터 음성 가입자의 수, 출력 link용량, 음성의 queue크기, 음성의 overflow확률에서는 서로 trade-off가 있고 data에서도 비슷한 tradeoff가 있음을 알았다. 또한 입력 traffic량과 link의 용량에 따라서 음성과 데이타간의 성능에서 서로 tradeoff가 있고, TASI의 이득이 2이상이고 음성가입자의 수가 적을 경우 데이타의 평균 지연시간은 buffer의 최대길이 보다 길음을 알아내었다.