Hong-Gi Yeom

ERS and ERD analysis during the imaginary movement of arms

Brain-Computer Interface (BCI) is an interface technique between human and computer which can help severely motor-disabled persons to communicate and control their environment. Many researchers use EEG signals for BCI. However, these studies are early stages and it has some problems such as a lack of accurate analysis and measurements with many electrodes. Therefore, we propose an EEG signals measurement method and analysis methods for BCI. Our purpose of this study is to recognize subjectpsilas intention when they move their arms. EEG signals are recorded during the imaginary movement of subjectpsilas arms at electrode positions Fp1, Fp2, C3, and C4. We analyzed ERS (Event-Related Synchronization) and ERD (Event-Related Desynchronization) which are detected when people move their limbs in the mu wave and beta wave. Results of this study showed that ERD occurred in mu waves and ERS occurred in beta waves at C3 during the imaginary movement of right arm. Similarly, ERD occurred in mu waves and ERS occurred in beta waves at C4 during the imaginary movement of left arm. And ERD occurred at Fp1 and Fp2, in both case.

Variance considered machines: Modification of optimal hyperplanes in support vector machines

We propose a new classification algorithm, variance considered machine (VCM), by modifying optimal hyperplanes of the support vector machine (SVM). The SVM is a good method to calculate a slope of optimal hyperplanes with maximal margin. However, this algorithm neglects to consider variances and prior probabilities of the data. It can increase probabilities of error. To solve this problem, the VCM shifts the optimal hyperplanes of the SVM according to variances and prior probabilities. Therefore, the VCM has not only maximal margin, which is an advantage of the SVM, but also lower error probability. Through 10 case examples with different variances and prior probabilities, we demonstrated the superiority of the VCM by comparing the results of the SVM and VCM.

Human Emotion Recognition using Power Spectrum of EEG Signals : Application of Bayesian Networks and Relative Power Values

Many researchers are studying about human Brain-Computer Interface(BCI) that it based on electroencephalogram(EEG) signals of multichannel. The researches of EEG signals are used for detection of a seizure or a epilepsy and as a lie detector. The researches about an interface between Brain and Computer have been studied robots control and game of using human brain as engineering recently. Especially, a field of brain studies used EEG signals is put emphasis on EEG artifacts elimination for correct signals. In this paper, we measure EEG signals as human emotions and divide it into five frequence parts. They are calculated related the percentage of selecting range to total range. the calculating values are compared standard values by Bayesian Network. lastly, we show the human face avatar as human Emotion.

Performance Improvements of Brain-Computer Interface Systems based on Variance-Considered Machines

This paper showed the possibilities of performance improvement of Brain-Computer Interface (BCI) decreasing classification error rates of EEG signals by applying Variance-Considered Machine (VCM) which proposed in our previous study. BCI means controlling system such as computer by brain signals. There are many factors which affect performances of BCI. In this paper, we used suggested algorithm as a classification algorithm, the most important factor of the system, and showed the increased correct rates. For the experiments, we used data which are measured during imaginary movements of left hand and foot. The results indicated that superiority of VCM by comparing error rates of the VCM and SVM. We had shown excellence of VCM with theoretical results and simulation results. In this study, superiority of VCM is demonstrated by error rates of real data.

Brain-Computer Interface based on Changes of EEG on Broca`s Area

In this paper, we measured EEG signals on frontal and Broca`s area when subjects imagine to speak A or B or C or D. These signals were analyzed by Event-Related Spectral Perturbation (ERSP), Inter-Trial Coherence (ITC) and Event Related Potential (ERP) methods. As a result, high coherences were showed at 113Hz during 0300ms after the stimuli of each character and P300 was seen clearly and there are several differences between the ERP results. However, unlike the motivation of this study to classify the characters, it is impossible that we can classify each intention or each character cause these differences. Nevertheless, this paper suggest an application system using this results so BCI can provide various services.

EEG Signals Measurement and Analysis Method for Brain-Computer Interface

There are many methods for Human-Computer Interface. Recently, many researchers are studying about Brain-Signal this is because not only the disabled can use a computer by their thought without their limbs but also it is convenient to general people. But, studies about it are early stages. This paper proposes an EEG signals measurement and analysis methods for Brain-Computer Interface. Our purpose of this research is recognition of subject`s intention when they imagine moving their arms. EEG signals are recorded during imaginary movement of subject`s arms at electrode positions Fp1, Fp2, C3, C4. We made an analysis ERS(Event-Related Synchronization) and ERD(Event-Related Desynchronization) which are detected when people move their limbs in the waves and waves. Results of this research showed that waves are decreased and waves are increased at left brain during the imaginary movement of right hand. In contrast, waves are decreased and waves are increased at right brain during the imaginary movement of left hand.

Soundsource Localization and Tracking System of Intruder for Intelligent Surveillance System

In the place that its security is crucial, the necessity of system which can tract and recognize random person is getting more important. In this paper, we`d like to develop the invader tracking system which consists of the sound source tracking-sensor and the pan-tilt camera for wide-area guard. After detecting the direction of any sound with the sound source tracking-sensor at first, our system make move the pan-tilt camera to that direction and extract reference image from that camera. This reference image is compared and updated by the next captured image after some interval time. By keeping on it over again, we can realize the guard system which can tract an invader using the difference image and the result of another image processing. By linking home network security system, the suggested system can provide some interfacing functions for the security service of the public facilities as well as that of home.

Membership Function-based Classification Algorithms for Stability improvements of BCI Systems

To improve system performance, we apply the concept of membership function to Variance Considered Machines (VCMs) which is a modified algorithm of Support Vector Machines (SVMs) proposed in our previous studies. Many classification algorithms separate nonlinear data well. However, existing algorithms have ignored the fact that probabilities of error are very high in the data-mixed area. Therefore, we make our algorithm ignore data which has high error probabilities and consider data importantly which has low error probabilities to generate system output according to the probabilities of error. To get membership function, we calculate sigmoid function from the dataset by considering means and variances. After computation, this membership function is applied to the VCMs.

Emotion Recognition and Expression System of User using Multi-Modal Sensor Fusion Algorithm

As they have more and more intelligence robots or computers these days, so the interaction between intelligence robot(computer) - human is getting more and more important also the emotion recognition and expression are indispensable for interaction between intelligence robot(computer) - human. In this paper, firstly we extract emotional features at speech signal and facial image. Secondly we apply both BL(Bayesian Learning) and PCA(Principal Component Analysis), lastly we classify five emotions patterns(normal, happy, anger, surprise and sad) also, we experiment with decision fusion and feature fusion to enhance emotion recognition rate. The decision fusion method experiment on emotion recognition that result values of each recognition system apply Fuzzy membership function and the feature fusion method selects superior features through SFS(Sequential Forward Selection) method and superior features are applied to Neural Networks based on MLP(Multi Layer Perceptron) for classifying five emotions patterns. and recognized result apply to 2D facial shape for express emotion.