Showing papers in "Artificial Life and Robotics in 2011"

Extraction of the minimum number of Gabor wavelet parameters for the recognition of natural facial expressions

Abstract: Facial expression recognition has recently become an important research area, and many efforts have been made in facial feature extraction and its classification to improve face recognition systems. Most researchers adopt a posed facial expression database in their experiments, but in a real-life situation the facial expressions may not be very obvious. This article describes the extraction of the minimum number of Gabor wavelet parameters for the recognition of natural facial expressions. The objective of our research was to investigate the performance of a facial expression recognition system with a minimum number of features of the Gabor wavelet. In this research, principal component analysis (PCA) is employed to compress the Gabor features. We also discuss the selection of the minimum number of Gabor features that will perform the best in a recognition task employing a multiclass support vector machine (SVM) classifier. The performance of facial expression recognition using our approach is compared with those obtained previously by other researchers using other approaches. Experimental results showed that our proposed technique is successful in recognizing natural facial expressions by using a small number of Gabor features with an 81.7% recognition rate. In addition, we identify the relationship between the human vision and computer vision in recognizing natural facial expressions.

Development of a pulse control-type MEMS microrobot with a hardware neural network

Abstract: This article presents the micro-electro-mechanical systems (MEMS) microrobot which demonstrates locomotion controlled by hardware neural networks (HNN). The size of the microrobot fabricated by the MEMS technology is 4 × 4 × 3.5 mm. The frame of the robot is made of silicon wafer, and it is equipped with a rotary-type actuator, a link mechanism, and six legs. The rotary-type actuator generates rotational movement by applying an electrical current to artificial muscle wires. The locomotion of the microrobot is obtained by the rotation of the rotary-type actuator. As in a living organism, the HNN realized robot control without using any software programs, A/D converters, or additional driving circuits. A central pattern generator (CPG) model was implemented as an HNN system to emulate the locomotion pattern. The MEMS microrobot emulated the locomotion method and the neural networks of an insect with the rotary-type actuator, the link mechanism, and the HNN. The microrobot performed forward and backward locomotion, and also changed direction by inputting an external trigger pulse. The locomotion speed was 0.325 mm/s and the step width was 1.3 mm.

Multi-robot-based intelligent security system

Abstract: This article describes a multiple security module-based intelligent security system that has multiple communication interfaces which can be applied in home automation. The interfaces of the intelligent security system contain wired RS485, wireless RF, and Internet. The detection modules of the system have both active and passive security modules. The passive security modules contain wired security modules and wireless security modules. The control unit of all security modules is a HOLTEK microchip. Each security module has two different interfaces. They use voice modules to alarm users of an event, and to transmit real-time event signals to the supervising computer via the wired RS485 or wireless RF interface. If an event occurs, the supervising computer calculates its belief values using Dempster-Shafter evidence theory according to the passive wired and wireless security modules. If the belief value is over a set threshold, the supervising computer commands the mobile robot to move to the event location, and receives a signal from the mobile robot via the wireless RF interface. The supervising computer recognizes the final decision output using Dempster-Shafter evidence theory, and displays the detection and decision output values on the monitor of the user interface. Finally, we present some experimental results using wired passive security modules, wireless passive security modules, and active security modules for fire detection and gas leakage detection using the experimental platform of the intelligent security system.

Swing-up and LQR stabilization of a rotary inverted pendulum

Abstract: In this article, we consider the swing-up and linear quadratic regulator (LQR) stabilization of a rotary inverted pendulum. A DC motor rotates a rigid arm. At the end of the rigid arm is a joint with a pendulum suspended from it. Two encoders check the degree of the rigid arm and the pendulum every 0.5 ms. This article describes a modified bang-bang control which swings the pendulum up safely and fast. In order to solve the stabilization problem, we used a linear quadratic regulator. When the user gives a large disturbance to the pendulum so that it loses its position, it quickly recovers to the upright position. Experimental results showed that the proposed bang-bang and LQR controllers can stabilize a rotary inverted pendulum system within 3.0 s from any starting point. The system also showed robustness to a large disturbance.

A playmate robot system for playing the rock-paper-scissors game with humans

Abstract: We have developed a playmate robot system for playing the rock-paper-scissors game with humans. The playmate robot recognizes the hand motions of a human using image processing without attaching any additional units to the human. The playmate robot system consists of three parts: a game management part, a hand motion recognition part, and a robot hand control part. The system functions as follows. (1) Before the game is played, the game management part decides on the motion of the robot hand from amongst rock, paper, and scissors. After the game is played, the robot develops a reaction using speech and facial expressions depending on the result of the game. (2) The hand motion recognition part recognizes the hand motion of the human. It does not use any additional units on the human’s body, only a camera on the robot. (3) The robot hand control part shows the motion of the robot hand. A robot hand has four fingers that are controlled independently. We have played the rock-paper-scissors game with this playmate robot.

A two-variable silicon neuron circuit based on the Izhikevich model

Abstract: The silicon neuron is an analog electronic circuit that reproduces the dynamics of a neuron. It is a useful element for artificial neural networks that work in real time. Silicon neuron circuits have to be simple, and at the same time they must be able to realize rich neuronal dynamics in order to reproduce the various activities of neural networks with compact, low-power consumption, and an easy-to-configure circuit. We have been developing a silicon neuron circuit based on the Izhikevich model, which has rich dynamics in spite of its simplicity. In our previous work, we proposed a simple silicon neuron circuit with low power consumption by reconstructing the mathematical structure in the Izhikevich model using an analog electronic circuit. In this article, we propose an improved circuit in which all of the MOSFETs are operated in the sub-threshold region.

A network visualization of stable matching in the stable marriage problem

Abstract: The stable marriage problem (SMP) seeks matchings between n women and n men which would result in stability, and not lead to divorce or extramarital affairs. We have introduced a network consisting of nodes which represent matchings, and links between nodes which attain stability by exchanging a partner between two pairs. The network is depicted with nodes laid out to involve several coordinates which indicate either women's satisfaction or men's or both. With the network visualization, regularity and symmetry can be made conspicuous in specific instances of the SMP such as the Latin SMP.

Design of an android robot head for stage performances

Abstract: In this article, an android robot head is proposed for stage performances. As is well known, an android robot is a type of humanoid robot which is considered to be more like a human than many others types. An android robot has human-like joint structures and artificial skin, and so is the robot which is closest to a human in appearance. To date, several android robots have been developed, but most of them have been made for research purposes or exhibitions. In this article, attention is drawn to the more commercial value of an android robot, especially in the acting field. EveR-3, the android robot described here, has already been used in commercial plays in the theater, and through these it has been possible to learn which features of an android robot are necessary for it to function as an actor. A new 9-DOF head has been developed for stage performances. The DOF are reduced when larger motors are used to make exaggerated expressions, because exaggerated expressions are more important on the stage than detailed, complex expressions. LED lights are installed in both cheeks to emphasize emotional expressions by changes in color in the way that make-up is used to achieve a similar effect on human faces. From these trials, a new head which is more suitable for stage performances has been developed.

Cooperative swarm control for multiple mobile robots using only information from PSD sensors

Abstract: Recently, many studies on swarm robotics have been conducted in which the aim seems to be the realization of an ability to perform complex tasks by cooperating with each other. Future progress and concrete applications are expected. The objective of this study was to construct a cooperative swarm system by using multiple mobile robots. First, multiple mobile robots with six position-sensitive detector (PSD) sensors were designed. A PSD sensor is a type of photo sensor. A control system was considered to realize swarm behavior, such as that shown by Ligia exotica, by using only information from the PSD sensors. Experimental results showed interesting behavior among the multiple mobile robots, such as following, avoidance, and schooling. The controller of the schooling mode was designed based on subsumption architecture. The proposed system was demonstrated to high school students at OPEN CAMPUS 2010, held in Tokyo University of Science, Yamaguchi.

Facial expression recognition of a speaker using vowel judgment and thermal image processing

Abstract: We have previously developed a method for the recognition of the facial expression of a speaker. For facial expression recognition, we previously selected three images: (i) just before speaking, (ii) speaking the first vowel, and (iii) speaking the last vowel in an utterance. By using the speech recognition system named Julius, thermal static images are saved at the timed positions of just before speaking, and when just speaking the phonemes of the first and last vowels. To implement our method, we recorded three subjects who spoke 25 Japanese first names which provided all combinations of the first and last vowels. These recordings were used to prepare first the training data and then the test data. Julius sometimes makes a mistake in recognizing the first and/or last vowel (s). For example, /a/ for the first vowel is sometimes misrecognized as /i/. In the training data, we corrected this misrecognition. However, the correction cannot be carried out in the test data. In the implementation of our method, the facial expressions of the three subjects were distinguished with a mean accuracy of 79.8% when they exhibited one of the intentional facial expressions of "angry," "happy," "neutral," "sad," and "surprised." The mean accuracy of the speech recognition of vowels by Julius was 84.1%.

A nonholonomic control method for stabilizing an X4-AUV

Abstract: A nonholonomic control method is considered for stabilizing all attitudes and positions (x, y, or z) of an underactuated X4 autonomous underwater vehicle (AUV) with four thrusters and six degrees of freedom (DOF), in which the positions are stabilized according to the Lyapunov stability theory. A dynamic model is first derived, and then a sequential nonlinear control strategy is implemented for the X4-AUV which is composed of translational and rotational subsystems. A controller for the translational subsystem stabilizes one position out of the x-, y-, and z-coordinates, whereas controllers for the rotational subsystems generate the desired roll, pitch, and yaw angles. Thus, the rotational controllers stabilize all the attitudes of the X4-AUV at the desired (x-, y-, or z-) position of the vehicle. Some numerical simulations are conducted to demonstrate the effectiveness of the proposed controllers.

An efficient identification scheme for a nonlinear polynomial NARX model

Abstract: Nonlinear polynomial NARX (nonlinear autoregressive with exogenous inputs) model identification often faces the problem of the huge size of the candidate pool, which makes the "wrapper" structure selection algorithm worked at low efficiency. In this article, a correlation-based orthogonal forward selection (COFS) algorithm is proposed to select the necessary input variables so that the candidate pool thus formed becomes tractable. What is more, it is trunked by an importance index-based term selection method, where a multiobjective evolutionary algorithm (MOEA)-based structure selection algorithm could be used to identify the polynomial model efficiently. Two numerical simulations are carried out to show the effectiveness of the proposed method.

Prediction of the electron flux environment in geosynchronous orbit using a neural network technique

Abstract: In this study, a neural network technique is adopted to predict the electron flux in a geosynchronous orbit using several items of solar wind data obtained by ACE spacecraft and magnetic variations observed on the ground as input parameters. Parameter tuning for the back-propagation learning method is attempted for the feed-forward neural network. As a result, the prediction using the combined data of solar wind and ground magnetic data shows a highest prediction efficiency of 0.61, which is enough to adapt to the actual use of the space environment prediction.

Development and experimental study of a novel pruning robot

Abstract: This article presents the development of a timberjack-like pruning robot. The climbing principal is an imitation of the climbing approach of timberjacks in Japan. The robot's main features include having its center of mass outside the tree, and an innovative climbing strategy fusing straight and spiral climbs. This novel design brings both lightweight and high climbing speed features to the pruning robot. We report our progress in developing the robot, focusing on straight climbing, behavior on uneven surfaces, and pruning.

Motion planning for a landmine-detection robot

Abstract: This article describes a landmine-detection system that contains a landmine-detection mobile robot and a following mobile robot. In this system, the landmine-detection robot goes ahead, and uses a landmine detector and a GPS module to find a landmine, records the coordinates of its location, and transmits these coordinates to the following mobile robot via a wireless RF interface. The following robot can record the location and orientation of the landmine-detection robot and all the landmines in the region. The following robot moves close to the landmine, and programs a path to avoid obstacles and landmines automatically. The driving system of the landmine-detection mobile robot uses a microprocessor dsPIC 30F4011 as the core, and controls two DC servomotors to program the motion path. The user interface of the landmine-detection robot and the following robot uses Borland C++ Builder language to receive the location data. In the experimental results, the landmine-detection robot records the location of landmines using a GPS module, and transmits the locations to the following robot via a wireless RF interface. The following robot avoids the landmines, and improves the safety of people or materials being carried through the landmine area.

Medical image diagnosis of lung cancer by a revised GMDH-type neural network using various kinds of neuron

Abstract: A revised group method of data handling (GMDH)-type neural network algorithm using various kinds of neuron is applied to the medical image diagnosis of lung cancer. The optimum neural network architecture for medical image diagnosis is automatically organized using a revised GMDH-type neural network algorithm, and the regions of lung cancer are recognized and extracted accurately. In this revised GMDH-type neural network algorithm, polynomial-type and radial basis function (RBF)-type neurons are used for organizing the neural network architecture in order to fit the complexity of the nonlinear system.

Developing a monitoring psychological stress index system via photoplethysmography

Abstract: The purpose of this work was to assess the psychological stress index (PSI) by using oxyhemoglobin saturation by SpO2 [i.e., the plethysmograph (PPG) signal], which can be measured easily and conveniently. We use the plethysmograph amplitude (PPGA) and the heart-beat interval (HBI) extracted from the SpO2 waveform to obtain the stress index and to quantify it from 0 to 100. The respiration rate can also be extracted from the heart-rate interval, as shown in our previous research. Therefore, the PPG signal can display the heart rate, SpO2 waveform, SpO2 concentration, respiration rate, and psychological stress index. This methodology was tested on nine volunteer Taiwanese students under the stress of an English presentation. The experiment, which takes 30 min for each student, was separated into three stages: preparation, presentation, and discussion, and relaxation. The PSI values of these three stages were 49 ± 10, 60 ± 11, and 56 ± 10, respectively. The results were very successful, and showed PSI values changing from low, to high, and back to low during these three stages. In the near future, we hope to implement this system into a robotic wheelchair in order to monitor the PSI of elderly patients in a nursing home, and to evaluate their psychological condition based on this method.

Super-twisting second-order sliding mode control for a synchronous reluctance motor

Abstract: This article presents the design and implementation of a super-twisting second-order sliding-mode controller (SOSMC) for a synchronous reluctance motor. Second-order sliding-mode control is an effective tool for the control of uncertain nonlinear systems since it overcomes the main drawbacks of the classical sliding-mode control, i.e., the large control effort and the chattering phenomenon. Its real implementation implies simple control laws, and ensures an improvement in the sliding accuracy with respect to conventional sliding-mode control. This article proposes a novel scheme that is based on the technique of super-twisting second-order sliding-mode control. First, SOSMC is derived mathematically, and then the performance of the proposed method is verified by simulations. The proposed SOSMC shows robustness for variations in the motor parameters and an improvement in the chattering phenomenon.

A system for facial expression recognition of a speaker using front-view face judgment, vowel judgment, and thermal image processing

Abstract: For facial expression recognition, we selected three images: (i) just before speaking, (ii) speaking the first vowel, and (iii) speaking the last vowel in an utterance. In this study, as a pre-processing module, we added a judgment function to distinguish a front-view face for facial expression recognition. A frame of the front-view face in a dynamic image is selected by estimating the face direction. The judgment function measures four feature parameters using thermal image processing, and selects the thermal images that have all the values of the feature parameters within limited ranges which were decided on the basis of training thermal images of front-view faces. As an initial investigation, we adopted the utterance of the Japanese name "Taro," which is semantically neutral. The mean judgment accuracy of the front-view face was 99.5% for six subjects who changed their face direction freely. Using the proposed method, the facial expressions of six subjects were distinguishable with 84.0% accuracy when they exhibited one of the intentional facial expressions of "angry," "happy," "neutral," "sad," and "surprised." We expect the proposed method to be applicable for recognizing facial expressions in daily conversation.

Outdoor autonomous navigation using SURF features

Abstract: In this article, we propose a speeded-up robust features (SURF)-based approach for outdoor autonomous navigation. In this approach, we capture environmental images using an omni-directional camera and extract features of these images using SURF. We treat these features as landmarks to estimate a robot's self-location and direction of motion. SURF features are invariant under scale changes and rotation, and are robust under image noise, changes in light conditions, and changes of viewpoint. Therefore, SURF features are appropriate for the self-location estimation and navigation of a robot. The mobile robot navigation method consists of two modes, the teaching mode and the navigation mode. In the teaching mode, we teach a navigation course. In the navigation mode, the mobile robot navigates along the teaching course autonomously. In our experiment, the outdoor teaching course was about 150 m long, the average speed was 2.9 km/h, and the maximum trajectory error was 3.3 m. The processing time of SURF was several times shorter than that of scale-invariant feature transform (SIFT). Therefore, the navigation speed of the mobile robot was similar to the walking speed of a person.

Blood flow velocity waveforms in the middle cerebral artery at rest and during exercise

Abstract: Blood flow velocities in the middle cerebral artery (MCA) were measured under steady-state and incremental cycle exercises using a transcranial Doppler ultrasound velocimeter. The peak systolic velocity was found to rise markedly under exercise, while the end diastolic velocity tended to remain at the resting value. The relationship between peak systolic velocity and systolic blood pressure, and that between peak systolic velocity and heart rate were found to show a statistically significant correlation. The mean MCA blood velocity also showed a significant correlation with the mean arterial pressure and heart rate. The fluctuations of velocity and the resistance index were calculated in order to evaluate the hemodynamic load on the vessel wall; these also increased markedly under exercise. Such hemodynamic changes in activity might be important in understanding the genesis of vascular diseases, as well as the physiology of cerebral circulation.

An adaptive resolution hybrid binary-real coded genetic algorithm

Abstract: In genetic algorithms (GAs), is it better to use binary encoding schemes or floating point encoding schemes? In this article, we try to tackle this controversial question by proposing a novel algorithm that divides the computational power between two cooperative versions of GAs. These are a binary-coded GA (bGA) and a real-coded GA (rGA). The evolutionary search is primarily led by the bGA, which identifies promising regions in the search space, while the rGA increases the quality of the solutions obtained by conducting an exhaustive search throughout these regions. The resolution factor (R), which has a value that is increasingly adapted during the search, controls the interactions between the two versions. We conducted comparison experiments employing a typical benchmark function to prove the feasibility of the algorithm under the critical scenarios of increasing problem dimensions and decreasing precision power.

Development of a motion analysis system using acceleration sensors for tennis and its evaluations

Abstract: The term biological motion is often used by researchers studying the patterns of movement generated by living forms and in sports. We studied a pattern recognition system of motion in sport using biological motion data. Biological motion data are acquired using a 3D motion capture system. However, 3D motion capture systems are very expensive. In this article, a biological motion capture system was built using acceleration sensors. Our proposed system uses the technique of Gaussian fitting and regression analysis. We tested our proposed system in pattern recognition of outdoor tennis and its evaluations.

Prime number generation using memetic programming

Abstract: For centuries, the study of prime numbers has been regarded as a subject of pure mathematics in number theory. Recently, this vision has changed and the importance of prime numbers has increased rapidly, especially in information technology, e.g., public key cryptography algorithms, hash tables, and pseudo-random number generators. One of the most popular topics to attract attention is to find a formula that maps the set of natural numbers into the set of prime numbers. However, to date there is no known formula that produces all primes. In this article, we use a hybrid evolutionary algorithm, called the memetic programming (MP) algorithm, to generate mathematical formulas that produce distinct primes. Using the MP algorithm, we succeeded in discovering an interesting set of formulas that produce sets of distinct primes.

Automatic detection of the topographical distribution of EEG rhythms based on an iterative adjustment of the averaged reference potential

Abstract: Electroencephalogram (EEG) interpretation is important for the investigation of brain diseases. In this study, an automatic technique was developed to detect the topographical distribution of EEG rhythms. In order to obtain the topographical distribution, the amplitude of the EEG rhythm was analyzed based on the referential derivation where the reference potential was adjusted iteratively. The result of the automatic detection of the topographical distribution was helpful in highlighting the EEG rhythms of interest for automatic EEG interpretation. The technique developed has application significance for real clinics.

A learning method for dynamic Bayesian network structures using a multi-objective particle swarm optimizer

Abstract: In this article, we present a multi-objective discrete particle swarm optimizer (DPSO) for learning dynamic Bayesian network (DBN) structures. The proposed method introduces a hierarchical structure consisting of DPSOs and a multi-objective genetic algorithm (MOGA). Groups of DPSOs find effective DBN sub-network structures and a group of MOGAs find the whole of the DBN network structure. Through numerical simulations, the proposed method can find more effective DBN structures, and can obtain them faster than the conventional method.

A survey of genetic algorithm-based face recognition

Abstract: Traditionally, special objects can be detected and recognized by the template matching method, but the recognition speed has always been a problem. In addition, for recognition by a neural network, training the data is always time-consuming. In this article, the current method of genetic algorithm-based face recognition is summarized, and experiments for real-time use are described. The chromosomes generated by the genetic algorithm (GA) contain information (parameters) about the face, and genetic operators are used to detect and obtain the position of the face of interest in an image. Here, the parameters of the coordinates (x, y) of the center of the face, the rate of scale, and the angle of rotation θ, are encoded into the GA.

EMG control of a pneumatic 5-fingered hand using a Petri net

Abstract: This article presents the control method for a 5-fingered artificial hand using electromyography (EMG) signals. Our targeted artificial hand is driven by pneumatic actuators to reduce its weight, and we use ON/OFF solenoid valves instead of electro pneumatic regulators to simplify the control system. The pneumatic hand has 15 degrees of freedom, and it seems difficult to reproduce all the finger motions from the EMG signals only. Therefore, we describe typical hand motions using a Petri net, and control the finger motions efficiently based on this model. Each state of the Petri net indicates one step in the hand posture to complete the intended motion. Simultaneously, this state corresponds to the ON/OFF pattern of the 15 solenoid valves. This enables the operator to control the 5-fingered dexterous hand smoothly, transiting the state in the Petri net according to the EMG motion signals. We conducted an experiment to verify the validity of the proposed method. In the experiment, five typical motions (spherical grasp, power grip, hook grip, key grip, precision grip) were successfully performed using the 6-channel EMG signals measured from the operator's forearm.

The study of path error for an omnidirectional home care mobile robot

Abstract: The first objective of this research was to develop an omnidirectional home care mobile robot. A PC-based controller controls the mobile robot platform. This service mobile robot is equipped with an “indoor positioning system” and an obstacle avoidance system. The indoor positioning system is used for rapid and precise positioning and guidance of the mobile robot. The obstacle avoidance system can detect static and dynamic obstacles. In order to understand the stability of a three-wheeled omnidirectional mobile robot, we carried out some experiments to measure the rectangular and circular path errors of the proposed mobile robot in this research. From the experimental results, we found that the path error was smaller with the guidance of the localization system. The mobile robot can also return to its starting point. The localization system can successfully maintain the robot’s heading angle along a circular path.

Trajectory control of biomimetic robots for demonstrating human arm movements

Abstract: This study describes the trajectory control of biomimetic robots by developing human arm trajectory planning. First, the minimum jerk trajectory of the joint angles is produced analytically, and the trajectory of the elbow joint angle is modified by a time-adjustment of the joint motion of the elbow relative to the shoulder. Next, experiments were conducted in which gyro sensors were utilized, and the trajectories observed were compared with those which had been produced. The results showed that the proposed trajectory control is an advantageous scheme for demonstrating human arm movements.