• The Korean Journal of Food And Nutrition
• /
• v.37 no.4
• /
• pp.197-210
• /
• 2024

This study aims to develop a deep learning model to monitor rice serving amounts in institutional foodservice, enhancing personalized nutrition management. The goal is to identify the best convolutional neural network (CNN) for detecting rice quantities on serving trays, addressing balanced dietary intake challenges. Both a vanilla CNN and 12 pre-trained CNNs were tested, using features extracted from images of varying rice quantities on white trays. Configurations included optimizers, image generation, dropout, feature extraction, and fine-tuning, with top-1 validation accuracy as the evaluation metric. The vanilla CNN achieved 60% top-1 validation accuracy, while pre-trained CNNs significantly improved performance, reaching up to 90% accuracy. MobileNetV2, suitable for mobile devices, achieved a minimum 76% accuracy. These results suggest the model can effectively monitor rice servings, with potential for improvement through ongoing data collection and training. This development represents a significant advancement in personalized nutrition management, with high validation accuracy indicating its potential utility in dietary management. Continuous improvement based on expanding datasets promises enhanced precision and reliability, contributing to better health outcomes.

• Journal of Korean Society of Forest Science
• /
• v.98 no.3
• /
• pp.351-356
• /
• 2009

This study purposed to judge potential possibility of building highly precise empirical model using environmental variables. Environmental variables such as altitude, mean annual rainfall, mean annual temperature and organic matter ratio of soil were added to height and diameter model for Chamaecyparis obtusa, and examined accuracy and residuals of prediction model. Improvement in precision was found for the Gompertz polymorphic height model by including mean temperature and altitude as independent variables, while the Gompertz diameter model with annual rainfall and altitude was showed improvement of precision and accuracy. Comparing the improvement of precision between the model before adding environmental variables and the model after adding them, an improvement or some ratio was obtained though it is not obvious. Therefore, there is enough proof that adding environmental variables, which can be easily acquired relatively when considering the difficulties of measurement and budget, into the model as independent variables would improve the accuracy and precision of growth models.

• Journal of Information Processing Systems
• /
• v.16 no.4
• /
• pp.809-819
• /
• 2020

Gender prediction accuracy increases as convolutional neural network (CNN) architecture evolves. This paper compares voting and ensemble schemes to utilize the already trained five CNN models to further improve gender prediction accuracy. The majority voting usually requires odd-numbered models while the proposed softmax-based voting can utilize any number of models to improve accuracy. The ensemble of CNN models combined with one more fully-connected layer requires further tuning or training of the models combined. With experiments, it is observed that the voting or ensemble of CNN models leads to further improvement of gender prediction accuracy and that especially softmax-based voters always show better gender prediction accuracy than majority voters. Also, compared with softmax-based voters, ensemble models show a slightly better or similar accuracy with added training of the combined CNN models. Softmax-based voting can be a fast and efficient way to get better accuracy without further training since the selection of the top accuracy models among available CNN pre-trained models usually leads to similar accuracy to that of the corresponding ensemble models.

• Journal of rehabilitation welfare engineering & assistive technology
• /
• v.10 no.3
• /
• pp.221-228
• /
• 2016

In this paper, we introduce our simulation software for EEG signal accuracy improvement. Users can check and train own EEG signal accuracy using our simulation software. Subjects were shown emotional imagination condition with landscape photography and logical imagination condition with a mathematical problem to subject. We use that EEG signal data, and apply Independent Component Analysis algorithm for noise removal. So we can have beta waves(${\beta}$, 14-30Hz) data through Band Pass Filter. We extract feature using Root Mean Square algorithm and That features are classified through Support Vector Machine. The classification result is 78.21% before EEG signal accuracy improvement training. but after successive training, the result is 91.67%. So user can improve own EEG signal accuracy using our simulation software. And we are expecting efficient use of BCI system based EEG signal.

• Journal of Institute of Control, Robotics and Systems
• /
• v.6 no.12
• /
• pp.1093-1098
• /
• 2000

This paper presents an efficient algorithm for including the kinematic calibration data into the motion controller to improve the positioning accuracy of the manipulators. Rather than spending several iterations for finding the inverse solution of the calibrated kinematics, our approach requires only the nominal inverse solution and the calibrated forward kinematics for providing a better position command promptly. Thus, real-time application is guaranteed whenever the manipulators nominal inverse solution can be expressed in a closed form. Experimental results show that the line tracking performances can be remarkably improved by employing our algorithm.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2001.10a
• /
• pp.263-268
• /
• 2001

Two ways of improvement of the strut for the HCF was made in this paper. The strut for the existing HCF is a passive link, which results the posture of the HCF and related bones. The accuracy of the HCF depends on the accuracy of the strut length. A "digital gage" was proposed to increase the accuracy of the strut by presenting the measuring result as figures in manual mode operation. And a "motorized strut system" was designed for the automated HCF operation. A strut was equipped with a motor, "motorized strut" can be operated manually and automatically. In automatic mode, the HCF operating data is generated by the HCF schedule package in PC and is downloaded to the "motorized strut system" controller. By these two improvements, changes in orthopedic equipments like HCF and other Ilizarov fixators are expected.

• Journal of the Korean Professional Engineers Association
• /
• v.22 no.3
• /
• pp.5-13
• /
• 1989

A major problem of stopwatch time study is how to do for the accurate and consistent performance rating, which is one of the critical variables to determine the accuracy of work measurement and should be still dependent upon time observer's judgement. Therefore the time observer's ability for the performance rating is very important, and must be improved by correct training method and procedure. This paper developed a new benchmark and benchmark-observation method for the effective performance rating training of assembling and machining operations. The trainees' ability in the accuracy and consistency of the performance rating ,improved significantly after being trained by subject method. The percentage improvement in rating accuracy and consistency values was 34.7% and 49% respectively. In addition, benchmark-practice method for the performance rating training is not significant, so it is proofed that the skill of a certain operation is not important for the improvement of the rating ability.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1997.10a
• /
• pp.393-396
• /
• 1997

High-speed machining is one of the most effective technologies to improve productivity. Because of the high speed and high feed rate, high-speed machining can give great advantages for the machining of dies and molds. This paper describes on the improvement of machining accuracy in high-speed machining. Depth of cut and feed rate are control factors. The effect of the control factors on machining accuracy is investigated using two-way factorial design.

• Journal of computational fluids engineering
• /
• v.21 no.1
• /
• pp.94-102
• /
• 2016

The present paper deals with accuracy improvement of a bleed boundary condition model used to improve the performance of supersonic inlets. In order to accurately predict the amount of bleed mass flow rates, this study performs a scaling of sonic flow coefficient data for 90-degree bleed holes in consideration of Prandtl-Meyer expansion theory. Furthermore, it is assumed that porosity varies with stream-wise location of the porous bleed plate to accurately predict downstream boundary layer profiles. The bleed boundary condition model is demonstrated through Computational Fluid Dynamics(CFD) simulations of bleed flows on a flat plate with/without an oblique shock. As a result, the bleed model shows the improved accuracy of bleed mass rates and downstream boundary layer profiles.

• Journal of the Korean Society for Precision Engineering
• /
• v.19 no.7
• /
• pp.88-96
• /
• 2002

High-speed machining is one of the most effective technologies to improve productivity. Because of the high speed and high feed rate, high-speed machining can give great advantages for the machining of dies and molds. This paper describes on the improvement of machining accuracy in high-speed machining. Depth of cut, feed rate and spindle revolution are control factors. The effect of the control factors on machining accuracy is investigated using two-way factorial design.