• Journal of KIISE:Computing Practices and Letters
• /
• v.11 no.3
• /
• pp.284-291
• /
• 2005

I propose a framework for detecting, recognizing and classifying facial features based on learned expression patterns. The framework recognizes facial expressions by using PCA and expression HMM(EHMM) which is Hidden Markov Model (HMM) approach to represent the spatial information and the temporal dynamics of the time varying visual expression patterns. Because the low level spatial feature extraction is fused with the temporal analysis, a unified spatio-temporal approach of HMM to common detection, tracking and classification problems is effective. The proposed recognition framework is accomplished by applying posterior probability between current visual observations and previous visual evidences. Consequently, the framework shows accurate and robust results of recognition on as well simple expressions as basic 6 facial feature patterns. The method allows us to perform a set of important tasks such as facial-expression recognition, HCI and key-frame extraction.

• Proceedings of the KIEE Conference
• /
• 2004.07d
• /
• pp.2631-2633
• /
• 2004

This paper describes the construction of a half sphere screen driving tilting simulator that can perform six degree-of-freedom( DOF) motions simmulator to a tilting train. The mathematical equations of Tilting Train dynamics are first derived from the 6-DOF bicycle model and incorporated with the bogie. carbody, and suspension subsystems. The equations of motion are then programmed by visual C++ code. To achieve the simulator functions. a motion platform that is constructed by six electric-driven actuators is designed. and its kinetics/inverse kinetics analysis is also conducted. Driver operation signals such as carbady angle, accelerator. and tilting positions are measured to trigger the Tilting dynamics calculation and further actuate the cylinders by the motion platform control program. In addition. a digital PID controller is added to achieve the stable and accurate displacements of the motion platform. The experiments prove that the designed simulator is adequate in performing some special rail mad driving situations discussed in this paper.

• The Bulletin of The Korean Astronomical Society
• /
• v.44 no.1
• /
• pp.52.1-52.1
• /
• 2019

The Helioseismic and Magnetic Imager (HMI) is the instrument of Solar Dynamics Observatory (SDO) to study the magnetic field and oscillation at the solar surface. The HMI image is not enough to analyze very small magnetic features on solar surface since it has a spatial resolution of one arcsec. Super resolution is a technique that enhances the resolution of a low resolution image. In this study, we use a method for enhancing the solar image resolution using a Deep-learning model which generates a high resolution HMI image from a low resolution HMI image (4 by 4 binning). Deep learning networks try to find the hidden equation between low resolution image and high resolution image from given input and the corresponding output image. In this study, we trained a model based on a very deep residual channel attention networks (RCAN) with HMI images in 2014 and test it with HMI images in 2015. We find that the model achieves high quality results in view of both visual and measures: 31.40 peak signal-to-noise ratio(PSNR), Correlation Coefficient (0.96), Root mean square error (RMSE) is 0.004. This result is much better than the conventional bi-cubic interpolation. We will apply this model to full-resolution SDO/HMI and GST magnetograms.

• Journal of Biosystems Engineering
• /
• v.27 no.3
• /
• pp.185-194
• /
• 2002

The objective of this study was to analyze a stability of sideways overturning of a forwarder developed by the Forestry Research Institute. The stability analysis was conducted using a multibody dynamic analysis program. VisualNastran Desktop. A solid model of the forwarder was constructed and its physical properties such as mass, mass center and mass moment of inertia were determined on 3D CAD modeler of the Solid Edge 8.0. 3D simulations of sideways overturning of the forwarder were also performed on the Visual Nastran using the solid model when it traveled across the slope and traversed over an up-hill side obstacle. Stability comparison between a bogie-wheeled and a 6-wheeled forwarders was also made and found that the bogie-wheeled forwarder was more stable than the 6-wheeled one on slopes. The safety speeds of the forwarder predicted by the simulation under various conditions can be used as a guideline for safe operations of forwarders in mountainous area. The technique of using a solid model for the sability analysis can also be applied successfully to other vehicles like agricultural tractors, loaders and construction equipments.

• Nuclear Engineering and Technology
• /
• v.41 no.3
• /
• pp.247-270
• /
• 2009

This paper reviews studies on information searching behavior in process control systems and discusses some implications learned from previous studies for use in human factors studies on nuclear power plants (NPPs) main control rooms (MCRs). Information searching behavior in NPPs depends on expectancy, value, salience, and effort. The first quantitative scanning model developed by Senders for instrument panel monitoring considered bandwidth (change rate) of instruments as a determining factor in scanning behavior. Senders' model was subsequently elaborated by other researchers to account for value in addition to bandwidth. There is also another type of model based on the operator's situation awareness (SA) which has been developed for NPP application. In these SA-based models, situation-event relations or rules on system dynamics are considered the most significant factor forming expectancy. From the review of previous studies it is recommended that, for NPP application, (1) a set of symptomatic information sources including both changed and unchanged symptoms should be considered along with bandwidth as determining factors governing information searching (or visual sampling) behavior; (2) both data-driven monitoring and knowledge-driven monitoring should be considered and balanced in a systematic way; (3) sound models describing mechanisms of cognitive activities during information searching tasks should be developed so as to bridge studies on information searching behavior and design improvement in HMI; (4) the attention-situation awareness (A-SA) modeling approach should be recognized as a promising approach to be examined further; and (5) information displays should be expected to have totally different characteristics in advanced control rooms. Hence much attention should be devoted to information searching behavior including human-machine interface (HMI) design and human cognitive processes.

• Journal of the Korea Computer Graphics Society
• /
• v.26 no.3
• /
• pp.49-59
• /
• 2020

In this paper, we propose a speech animation synthesis specialized in Korean through a rule-based co-articulation model. Speech animation has been widely used in the cultural industry, such as movies, animations, and games that require natural and realistic motion. Because the technique for audio driven speech animation has been mainly developed for English, however, the animation results for domestic content are often visually very unnatural. For example, dubbing of a voice actor is played with no mouth motion at all or with an unsynchronized looping of simple mouth shapes at best. Although there are language-independent speech animation models, which are not specialized in Korean, they are yet to ensure the quality to be utilized in a domestic content production. Therefore, we propose a natural speech animation synthesis method that reflects the linguistic characteristics of Korean driven by an input audio and text. Reflecting the features that vowels mostly determine the mouth shape in Korean, a coarticulation model separating lips and the tongue has been defined to solve the previous problem of lip distortion and occasional missing of some phoneme characteristics. Our model also reflects the differences in prosodic features for improved dynamics in speech animation. Through user studies, we verify that the proposed model can synthesize natural speech animation.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.2
• /
• pp.693-709
• /
• 2018

In this paper, we propose a compact Convolutional Neural Network (CNN)-based tracker in conjunction with a particle filter architecture, in which the CNN model operates as an accurate candidates estimator, while the particle filter predicts the target motion dynamics, lowering the overall number of calculations and refines the resulting target bounding box. Experiments were conducted on the Online Object Tracking Benchmark (OTB) [34] dataset and comparison analysis in respect to other state-of-art has been performed based on accuracy and precision, indicating that the proposed algorithm outperforms all state-of-the-art trackers included in the OTB dataset, specifically, TLD [16], MIL [1], SCM [36] and ASLA [15]. Also, a comprehensive speed performance analysis showed average frames per second (FPS) among the top-10 trackers from the OTB dataset [34].

• Transactions of the Korean Society of Automotive Engineers
• /
• v.14 no.3
• /
• pp.28-34
• /
• 2006

An ACC driving simulator is a virtual reality device which designed to test or evaluate vehicle control algorithm. It is designed and built based on the rapid control prototyping(RCP) concept. Therefore this simulator adopt RCP tools to solve the equation of a vehicle dynamics model and control algorithm in real time, rendering engine to provide real-time visual representation of vehicle behavior and CAN communication to reduce networking load. It can provide also many different driving test environment and driving scenarios.

• Korean journal of applied entomology
• /
• v.51 no.3
• /
• pp.235-243
• /
• 2012

Population dynamics of the American serpentine leafminer, Liriomyza trifolii (Burgess), were observed and modeled in order to compare the effects of air and tomato leaf temperatures inside a greenhouse using DYMEX model builder and simulator (pre-programed module based simulation programs developed by CSIRO, Australia). The DYMEX model simulator consisted of a series of modules with the parameters of temperature dependent development and oviposition models of L. trifolii were incorporated from pre-published data. Leaf surface temperatures of cherry tomato leaves (cv. 'Koko') were monitored according to three tomato plant positions (top, > 1.8 m above the ground level; middle, 0.9 - 1.2 m; bottom, 0.3 - 0.5 m) using an infrared temperature gun. Air temperature was monitored at the same three positions using a self-contained temperature logger. Data sets for the observed air temperature and average leaf surface temperatures were collected (top and bottom surfaces), and incorporated into the DYMEX simulator in order to compare the effects of air and leaf surface temperature on the population dynamics of L. trifolii. The initial population consisted of 50 eggs, which were laid by five female L. trifolii in early June. The number of L. trifolii larvae was counted by visual inspection of the tomato plants in order to verify the performance of DYMEX simulation. The egg, pupa, and adult stage of L. trifolii could not be counted due to its infeasible of visual inspection. A significant positive correlation between the observed and the predicted numbers of larvae was found when the leaf surface temperatures were incorporated into the DYMEX simulation (r = 0.97, p < 0.01), but no significant positive correlation was observed with air temperatures(r = 0.40, p = 0.18). This study demonstrated that the population dynamics of L. trifolii was affected greatly by the leaf temperatures, though to little discernible degree by the air temperatures, and thus the leaf surface temperature should be for a consideration in the management of L. trifolii within cherry tomato greenhouses.

• Journal of Biosystems Engineering
• /
• v.39 no.4
• /
• pp.389-399
• /
• 2014

Purpose: This study was aimed at developing a tractor-driving simulator for the safety training of agricultural tractor operators. Methods: The developed simulator consists of five principal components: mock operator control devices, a data acquisition and processing device, a motion platform, a visual system that displays a computer model of the tractor, a motion platform, and a virtual environment. The control devices of a real tractor cabin were successfully converted into mock operator control devices in which sensors were used for relevant measurements. A 3D computer model of the tractor was also implemented using 3ds Max, tractor dynamics, and the physics of Unity 3D. The visual system consisted of two graphic cards and four monitors for the simultaneous display of the four different sides of a 3D object to the operator. The motion platform was designed with two rotational degrees of freedom to reduce cost, and inverse kinematics was used to calculate the required motor positions and to rotate the platform. The generated virtual environment consisted of roads, traffic signals, buildings, rice paddies, and fields. Results: The effectiveness of the simulator was evaluated by a performance test survey administered to 128 agricultural machinery instructors, 116 of whom considered the simulator as having potential for improving safety training. Conclusions: From the study results, it is concluded that the developed simulator can be effectively used for the safety training of agricultural tractor operators.