• Journal of the Ergonomics Society of Korea
• /
• v.30 no.6
• /
• pp.705-713
• /
• 2011

Objective: The aim of this study is to compare results of emotion recognition by several algorithms which classify three different emotional states(happiness, neutral, and surprise) using physiological features. Background: Recent emotion recognition studies have tried to detect human emotion by using physiological signals. It is important for emotion recognition to apply on human-computer interaction system for emotion detection. Method: 217 students participated in this experiment. While three kinds of emotional stimuli were presented to participants, ANS responses(EDA, SKT, ECG, RESP, and PPG) as physiological signals were measured in twice first one for 60 seconds as the baseline and 60 to 90 seconds during emotional states. The obtained signals from the session of the baseline and of the emotional states were equally analyzed for 30 seconds. Participants rated their own feelings to emotional stimuli on emotional assessment scale after presentation of emotional stimuli. The emotion classification was analyzed by Linear Discriminant Analysis(LDA, SPSS 15.0), Support Vector Machine (SVM), and Multilayer perceptron(MLP) using difference value which subtracts baseline from emotional state. Results: The emotional stimuli had 96% validity and 5.8 point efficiency on average. There were significant differences of ANS responses among three emotions by statistical analysis. The result of LDA showed that an accuracy of classification in three different emotions was 83.4%. And an accuracy of three emotions classification by SVM was 75.5% and 55.6% by MLP. Conclusion: This study confirmed that the three emotions can be better classified by LDA using various physiological features than SVM and MLP. Further study may need to get this result to get more stability and reliability, as comparing with the accuracy of emotions classification by using other algorithms. Application: This could help get better chances to recognize various human emotions by using physiological signals as well as be applied on human-computer interaction system for recognizing human emotions.

• Journal of IKEEE
• /
• v.20 no.3
• /
• pp.209-219
• /
• 2016

SVM is a machine learning method used for image processing. It is well known for its high classification performance. We have to perform multiple MAC operations in order to use SVM for image classification. However, if the resolution of the target image or the number of classification cases increases, the execution time of SVM also increases, which makes it difficult to be performed in real-time applications. In this paper, we propose an hardware architecture which enables real-time applications using SVM classification. We used parallel architecture to simultaneously calculate MAC operations, and also designed the system for several feature extractors for compatibility. RBF kernel was used for hardware implemenation, and the exponent calculation formular included in the kernel was modified to enable fixed point modelling. Experimental results for the system, when implemented in Xilinx ZC-706 evaluation board, show that it can process 60.46 fps for $1360{\times}800$ resolution at 100MHz clock frequency.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.5B
• /
• pp.469-479
• /
• 2009

The GEM frame is used a mean to deliver the variable length user data and consists of the header and the payload in the G-PON system. The HEC field of header protects contents of the header and is used to maintain GEM frame synchronization at the same time. When an LCDG (Loss of GEM Channel Delineation) occurs while receiving frames, the receiver have to discard corrupted frames until acquiring the synchronization again. Accordingly, high-speed synchronization method is required to minimize the frame loss. In this paper, we suggest not only a main state machine but a sub-state machine to reduce the frame loss when undetectable errors occurred in the GEM header. Also, we provide a more efficient and fast parallel structure to detect the starting point of the header. Finally, the proposed method is implemented with the FPGA and verified by the logic analyzer.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.19 no.1
• /
• pp.106-112
• /
• 2015

In this paper, we propose an on-line signature recognition algorithm using fingertip point in the air from the depth image acquired by Kinect. We use ten statistical features for each X, Y, Z axis to react to changes in Shifting and Scaling of the signature trajectories in three-dimensional space. Artificial Neural Network is a machine learning algorithm used as a tool to solve the complex classification problem in pattern recognition. We implement the proposed algorithm to actual on-line signature recognition system. In experiment, we verify the proposed method is successful to classify 4 different on-line signatures.

• Journal of the Korean institute of surface engineering
• /
• v.34 no.1
• /
• pp.62-67
• /
• 2001

Green technology is being developed up to a point that is feasible not only in an environmental sense, but also in an economical viewpoint. This paper introduces two case studies that applied $CO_2$ technology into nuclear industry. 1) Nuclear laundry : A laundry machine that uses liquid and supercritical $CO_2$ as a solvent for decontamination of contaminated working dresses in nuclear power plants was developed. The machine consists of a 16 liter reactor, a recovery system with compressors, and storage tanks. All $CO_2$ used in cleaning is fully recovered and reused in next cleaning, resulting in no production of secondary nuclear waste. Decontamination factor is still lower than that in the methods currently used in the plant. Nuclear laundry using $CO_2$ looks promising with technical improvements-surfactants and mechanical agitation. 2) $CO_2$ nozzle decontamination : An adjustable nozzle for controlling the size of dry ice snow was developed. Using the developed nozzle, a surface decontamination device was made. Human oils like fingerprints on glass were easy to remove. Decontamination ability was tested using a contaminated pump-housing surface. About 40 to 80% of radioactivity was removed. This device is effective in surface-decontamination of any electrical devices like detector, controllers which cannot be cleaned in aqueous solution.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.16 no.1
• /
• pp.11-19
• /
• 2011

This paper describe development of a hardware simulator for the DFIG wind power system, which was designed considering wind characteristic, blade characteristic, and blade inertia compensation. The simulator consists of three major parts, such as wind turbine model using induction motor, doubly-fed induction generator, converter-inverter set. and control system. The turbine simulator generates torque and speed signals for a specific wind turbine with respect to the given wind speed which is detected by Anemometer. This torque and speed signals are scaled down to fit the input of 3.5kW DFIG. The MSC operates to track the maximum power point, and the GSC controls the active and reactive power supplied to the grid. The operational feasibility was verified through computer simulations with PSCAD/EMTDC. And the implementation feasibility was confirmed through experimental works with a hardware set-up.

• Journal of radiological science and technology
• /
• v.11 no.2
• /
• pp.65-71
• /
• 1988

Radiation dose outside the radiotherapy treatment field can be significant and therefore is of clinical interest estimating organ dose. We have made measurements of dose at distances up to 70 cm from the central axis of $5{\times}5$, $10{\times}10$, $15{\times}15$, and $25{\times}25$ cm radiation fields of Co-60 ${\gamma}-ray$, at 5 cm depth in water. Contributions to the total secondary radiation dose from water scatter, machine (collimator) scatter and leakage radiation have been seperated. We have found that the component of dose from water scatter can be described by simple exponential function of distance from the central axis of the radiation field for all field sizes. Machine scatter contributes 20 to 60% of the total secondary dose depending on field size and distance from the field. Leakage radiation contributes very little dose, but becomes the dominant componant at distance beyond 40 cm from the central axis. Then, wedges can cause a factor 2 to 3 increase in dose at any point outside the field compared with the dose when no wedge is used. Adding blocks to a treatment field can cause an increase in dose at points outside the field, but the effect is much smaller than the effect of a wedge. From the results of these measurements, doses to selected organs outside the field for specified treatment geometries were estimated, and the potential for reducing these organ doses by additional shielding was assessed.

• Journal of the Korea Society of Computer and Information
• /
• v.24 no.11
• /
• pp.41-49
• /
• 2019

We are to build an unsupervised machine learning-based language model which can estimate the amount of information that are in need to process words consisting of subword-level morphemes and syllables. We are then to investigate whether the reading times of words reflecting their morphemic and syllabic structures are predicted by an information-theoretic measure such as surprisal. Specifically, the proposed Morfessor-based unsupervised machine learning model is first to be trained on the large dataset of sentences on Sejong Corpus and is then to be applied to estimate the information-theoretic measure on each word in the test data of Korean words. The reading times of the words in the test data are to be recruited from Korean Lexicon Project (KLP) Database. A comparison between the information-theoretic measures of the words in point and the corresponding reading times by using a linear mixed effect model reveals a reliable correlation between surprisal and reading time. We conclude that surprisal is positively related to the processing effort (i.e. reading time), confirming the surprisal hypothesis.

• Advances in aircraft and spacecraft science
• /
• v.3 no.3
• /
• pp.351-366
• /
• 2016

Macroperforations improve the sound absorption performance of porous materials in acoustic cavities and in waveguides. In an acoustic cavity, enhanced noise reduction is achieved using porous materials having macroperforations. Double porosity materials are obtained by filling these macroperforations with different poroelastic materials having distinct physical properties. The locations of macroperforations in porous layers can be chosen based on cavity mode shapes. In this paper, the effect of variation of macroporosity and double porosity in porous materials on noise reduction in an acoustic cavity is presented. This analysis is done keeping each perforation size constant. Macroporosity of a porous material is the fraction of area covered by macro holes over the entire porous layer. The number of macroperforations decides macroporosity value. The system under investigation is an acoustic cavity having a layer of poroelastic material rigidly attached on one side and excited by an internal point source. The overall sound pressure level (SPL) inside the cavity coupled with porous layer is calculated using mixed displacement-pressure finite element formulation based on Biot-Allard theory. A 32 node, cubic polynomial brick element is used for discretization of both the cavity and the porous layer. The overall SPL in the cavity lined with porous layer is calculated for various macroporosities ranging from 0.05 to 0.4. The results show that variation in macroporosity of the porous layer affects the overall SPL inside the cavity. This variation in macroporosity is based on the cavity mode shapes. The optimum range of macroporosities in poroelastic layer is determined from this analysis. Next, SPL is calculated considering periodic and nodal line based optimum macroporosity. The corresponding results show that locations of macroperforations based on mode shapes of the acoustic cavity yield better noise reduction compared to those based on nodal lines or periodic macroperforations in poroelastic material layer. Finally, the effectiveness of double porosity materials in terms of overall sound pressure level, compared to equivolume double layer poroelastic materials is investigated; for this the double porosity material is obtained by filling the macroperforations based on mode shapes of the acoustic cavity.

• Imaging Science in Dentistry
• /
• v.51 no.3
• /
• pp.299-306
• /
• 2021

Purpose: This study aimed to propose a fully automatic landmark identification model based on a deep learning algorithm using real clinical data and to verify its accuracy considering inter-examiner variability. Materials and Methods: In total, 950 lateral cephalometric images from Yonsei Dental Hospital were used. Two calibrated examiners manually identified the 13 most important landmarks to set as references. The proposed deep learning model has a 2-step structure-a region of interest machine and a detection machine-each consisting of 8 convolution layers, 5 pooling layers, and 2 fully connected layers. The distance errors of detection between 2 examiners were used as a clinically acceptable range for performance evaluation. Results: The 13 landmarks were automatically detected using the proposed model. Inter-examiner agreement for all landmarks indicated excellent reliability based on the 95% confidence interval. The average clinically acceptable range for all 13 landmarks was 1.24 mm. The mean radial error between the reference values assigned by 1 expert and the proposed model was 1.84 mm, exhibiting a successful detection rate of 36.1%. The A-point, the incisal tip of the maxillary and mandibular incisors, and ANS showed lower mean radial error than the calibrated expert variability. Conclusion: This experiment demonstrated that the proposed deep learning model can perform fully automatic identification of cephalometric landmarks and achieve better results than examiners for some landmarks. It is meaningful to consider between-examiner variability for clinical applicability when evaluating the performance of deep learning methods in cephalometric landmark identification.