Search

A motor imagery brain-computer interface system is used to translate a subject’s intention into a control command of machine, such as electrical wheelchair, robot manipulator, and so on. The overall process of classification of the motor imagery EEG signals is based on the acquisition of raw data from multiple channel of scalp when the subject tries to imagine the movement of limbs. So far, we have been concentrated which channel are activated by the imagination of the movement of limbs. Therefore, we have expected that the more channels are selected, the better results can be acquired. However, the problem is that using many channels causes other problems. When applying a common spatial pattern (CSP), which is a spatial feature extraction, many channels cause an overfitting problem, in addition there is difficulty using this technique for medical analysis. To overcome these problems, we suggest a binary particle swarm optimization (BPSO) as an optimal channel selection method. This paper examines selecting optimal channels and their combination, and comparing accuracy and the number of selected channels obtained from BPSO and simple genetic algorithm.

Emotion recognition is one of the important part to develop in human-human and human-computer interaction. In this paper, we focused on the experimental paradigm and feature extraction to extract features from the physiological signals. The experimental paradigm for data acquisition used MULTI module equipment of biofeedback 2000 x-pert which combined multi-sensor such as skin conductance, skin temperature, and blood volume pulse to collect physiological signals from the subject’s fingertip of the non-dominant hand. And an approach for the emotions recognition based on physiological signals such as fear, disgust, joy, and neutrality that international affective picture system (IAPS) was used to elicit emotion. These were selected to extract the characteristic parameters, which will be used for classifying emotions. Support vector machine (SVM) is a popular technique for classifying emotion recognition and perform high accuracy for classification. The experiment results showed that the methodology by using experimental paradigm, feature extraction and especially multi-class support vector machine (MSVM) provided significant improvement in accuracy for classification emotion recognition states.

Subject dependent nature of electroencelography (EEG) signal elicits in the imagination task cause the drop of accuracy of the classifier in the brain-computer interface when the system apply in different subjects or crossing session experiment. The main components that have effect most in this problem are spatial-spectral-temporal parameter of the EEG signal that need to extract to find the optimal solution in the BCI system. In this paper we proposed a method for extracting the optimal parameters based on particle swarm optimization algorithm. First EEG signals were enhanced by Laplace and band pass filter. Optimal spatio-spectral-temporal component of Principle Component Analysis were search by Particle Swarm Optimization (PSO) using Short Time Fourier Transform features and classification error rate from Support Vector Machine (SVM) as fitness function. With optimal parameters, principle component from the STFT features were extracted and combined into single optimal feature vector. 5 fold-cross validations are applied to SVM.

Brain-computer interface based on motor imagery is a system that transforms a subject’s intention into a control signal by classifying EEG signals obtained from the imagination of movement of a subject’s limbs. For the new paradigm, we do not know which positions are activated or not. A simple approach is to use as many channels as possible. Using many channels cause other problems. When applying a common spatial pattern (CSP), which is an EEG extraction method, many channels cause an overfitting problem, in addition there is difficulty using this technique for medical analysis. To overcome these problems, we suggest a particle swarm optimization applied to CSP. This paper examines selecting optimal channels among all channels, and comparing the classification accuracy between CSP and CSP with PSO by linear discriminant analysis.

Communication between machine and human is one of important application of brain-computer interface (BCI). There are still obtained many kinds of noise in the recorded human signals, specifically brain signal is electroencelophagram (EEG). It caused by outer and inner of the brain signals such as artifacts in signal, properties of EEG signal nonstationary, variant by time and subjects, which affect to the classification results. The most famous spatial filter in BCI context is common spatial patterns (CSP), maximize one condition while minimize the other condition using covariance. So in this experiment we recorded signal by using auditory stimuli to reduce artifact by gaze attention. Extended CSP methods were applied in this experiment to upgrade the classification accuracy of brain source separate by independent component analysis (ICA). We supposed this combination could purify the signals as 2 steps and increased the accuracy classification.

If we want to recognize the human's emotion via the face-to-face interaction, first of all, we need to extract the emotional features from the facial image and recognize the emotional states. Our facial emotional feature detection and extracting based on Active Appearance Models (AAM) with Ekman's Facial Action Coding System (FACS). Our approach to facial emotion recognition lies in the dynamic and probabilistic framework based on Dynamic Bayesian Network (DBN). The active appearance model (AAM) is a well-known method that can represent a non-rigid object, such as face, facial expression. In this paper, our approach to facial feature extraction lies in the proposed feature extraction method based on combining AAM with Facial Action Units of Facial Action Coding System (FACS) for automatically modeling and extracting the facial emotional features. Also, we use the Dynamic Bayesian Networks (DBNs) for modeling and understanding the temporal phases of facial expressions in image sequences.

This paper presents an original feature selection method for Emotion Recognition which includes many original elements. Feature selection has some merit regarding pattern recognition performance. Thus, we developed a method called an 'Interactive Feature Selection(IFS)' and 'GA Feature Selection(GAFS)'. Afterwards, the results (selected features) of the IFS and GAFS were applied to an emotion recognition system (ERS), which was also implemented in this research. Especially, our interactive feature selection method was based on a Reinforcement Learning Algorithm since it required responses from human users. By performing the IFS, we were able to obtain three top features and apply them to the ERS. We compared those results from a random selection and Sequential Forward Selection (SFS) and Genetic Algorithm Feature Selection (GAFS).

In this paper, we proposed the image processing techniques for extracting the cracks in a concrete surface crack image and the ART1-based RBF network for recognizing the directions of the extracted cracks. The image processing techniques used are the closing operation of morphological techniques, the Sobel masking used to extract edges of the cracks, and the iterated binarization for acquiring the binarized image from the crack image. The cracks are extracted from the concrete surface image after applying two times of noise reduction to the binarized image. We proposed the method for automatically recognizing the directions (horizontal, vertical, -45 degree, 45 direction degree) of the cracks with the ART1-based network. The proposed ART1-based RBF network applied ART1 to the learning between the input layer and the middle layer and the Delta learning method to the learning between the middle layer and the output layer. The experiments using real concrete crack images showed that the cracks in the concrete crack images were effectively extracted and the proposed ART1-based RBF network was effective in the recognition of the direction of extracted cracks.

This paper presents an original feature selection method for Emotion Recognition which includes many original elements. Feature selection has some merit regarding pattern recognition performance. Thus, we implemented a simulator called an 'IFS system' and the results of the IFS were applied to an emotion recognition system(ERS). Our innovative feature selection method was based on a Reinforcement Learning Algorithm and since it required responses from human users, it was denoted an 'Interactive Feature Selection'. By performing an IFS, we were able to obtain three top features and apply them to the ERS.

This paper presents an evolutionary cellular automata based neural systems (Evolutionary CANS) for visual servoing of RV-M2 robot manipulator. The architecture of CANS consist of a two-dimensional (2-D) array of basic neurons. Each neuron of CANS has local connections only with contiguous neuron and acts as a form of pulse according to the dynamics of the chaotic neuron model. CANS are generated from initial cells according to the cellular automata (CA) rule. Therefore neural architecture is determined by both initial pattern of cells and production rule of CA. Production rules of CA are evolved based on a DNA coding. DNA coding has the redundancy and overlapping of gene and is apt for representation of the rule. In this paper we show the general expression of CA rule and propose translating method from DNA code to CA rule. In addition, we present visual servoing application using evolutionary CANS.

Segmentation for the region of nucleus in the image of uterine cervical cytodiagnosis is known as the most difficult and important part in the automatic cervical cancer recognition system. In this paper, the region of nucleus is extracted from an image of uterine cervical cytodiagnosis using the HSI model. The characteristics of the nucleus are extracted from the analysis of morphemetric features, densitometric features, colorimetric features, and textural features based on the detected region of nucleus area. The classification criterion of a nucleus is defined according to the standard categories of the Bethesda system. The fuzzy ART algorithm is used to the extracted nucleus and the results show that the proposed method is efficient in nucleus recognition and uterine cervical Pap-Smears extraction.

In this paper, we propose an improved fuzzy RBF network which dynamically adjusts the rate of learning by applying the Delta-bar-Delta algorithm in order to improve the learning performance of fuzzy RBF networks. The proposed learning algorithm, which combines the fuzzy C-Means algorithm with the generalized delta learning method, improves its learning performance by dynamically adjusting the rate of learning. The adjustment of learning rate is achieved by self-generating middle-layered nodes and applying the Delta-bar-Delta algorithm to the generalized delta learning method for the learning of middle and output layers. To evaluate the learning performance of the proposed RBF network, we used 40 identifiers extracted from a container image as the training data. Our experimental results show that the proposed method consumes less training time and improves the convergence of learning, compared to the conventional ART2-based RBF network and fuzzy RBF network.

This paper presents the hexagon-based Q-leaning for object search with multiple robots. We set up an experimental environment with five small mobile robots, obstacles, and a target object. The robots were out to search for a target object while navigating in a hallway where some obstacles were placed. In this experiment, we used three control algorithms: a random search, an area-based action making (ABAM) process to determine the next action of the robots, and hexagon-based Q-learning to enhance the area-based action making process.

In this paper, we introduce a system that recognize emotion by speech and show the facial expression by using 2-dimensional emotion space. 4 emotional states are classified by the work with ANN. The derived features of the signal, pitch, and loudness are quantitatively contributed to the classification of emotions. Firstly we analyze the acoustic elements for using as emotional features and the elements are evaluated by ANN classifier. Secondly, we implement an avatar (simply drawn face) and the facial expressions are changed naturally by using the dynamic emotion space model.