Koji Yamada

University of the Ryukyus · Department of Information Engineering

When an electric wheelchair is operated using gaze motion, eye movements such as checking the environment and observing objects are also incorrectly recognized as input operations. This phenomenon is called the “Midas touch problem”, and classifying visual intentions is extremely important. In this paper, we develop a deep learning model that estim...

Recently, there have been attempts to use eye gaze as an interface for electric wheelchairs. In this study, we develop a VR simulator for a gaze-input type electric wheelchair and measure gaze data to analyze gaze behavior during electric wheelchair operation.

In recent years, various user interfaces have been developed to meet the diverse needs of physically disabled persons. In this paper, we developed a method to identify gazing and facial movements based on gaze time and eye/face information and developed an electric wheelchair that can be operated with the user's natural gazing and facial movements....

In recent years, various user interfaces have been developed to meet the diverse needs of physically disabled persons. In this paper, we developed a method to identify gazing and facial movements based on gaze time and eye/face information and developed an electric wheelchair that can be operated with the user’s "natural gazing and facial movements...

In this research, we examined an operation support method by gaze input using gaze amount histogram. In order to accurately identify the direction the user wishes to move and the direction which the attention is directed, in the gaze driving, estimation system based on the gaze amount histogram was constructed and a running experiment was performed...

Deep learning solves many classification problems. However, it is difficult to solve problems with imbalanced data. Therefore, the data volume is increased for the purpose of balancing. This is called data augmentation. Generally, the method of image data augmentation uses noise addition, rotation, and the like. Recently, images are generated using...

In this research, we examined an operation support method by gaze input using gaze amount histogram. In order to accurately identify the direction the user wishes to move and the direction which the attention is directed, in the gaze driving, estimation system based on the gaze amount histogram was constructed and a running experiment was performed...

In this study, we develop an automatic traveling system for an electric wheelchair with a status confirmation and obstacle avoidance function for realization of safe electric wheelchair running of the physically disabled persons in indoor. In this paper, we propose the automatic traveling system for an electric wheelchair and evaluate about accurac...

Facial expressions play an important role in communication as much as words. In facial expression recognition by human, it is difficult to uniquely judge, because facial expression has the sway of recognition by individual difference and subjective recognition. Therefore, it is difficult to evaluate the reliability of the result from recognition ac...

In nearby special support schools, the practice of driving a motorized wheelchair using a joystick has been conducted for physically disabled children. However, physically disabled children require support and adjustment of the equipment corresponding to their disabilities because it is difficult to operate a joystick in a specific direction owing...

The estimation methods for Twitter user’s attributes typically require a vast amount of labeled data. Therefore, an efficient way is to tag the unlabeled data and add it to the set. We applied the self-training SVM as a semisupervised method for age estimation and introduced Plat scaling as the unlabeled data selection criterion in the self-trainin...

In this study, we carried out the facial expression recognition from facial expression dataset using Convolutional Neural Networks (CNN). In addition, we analyzed intermediate outputs of CNN. As a result, we have obtained a emotion recognition score of about 58%; two emotions (Happiness, Surprise) recognition score was about 70%. We also confirmed...

In the nearby special support school, driving practice of motorized wheelchair intended for the physically disabled child have been conducted. We developed the joystick-type controller that adjusts the reaction angle automatically to suit the actual condition of the physically disabled child so far and it is lead to practice the driving of the moto...

As survival rate of cerebrovascular disease due to improvements in medical care, the number of patients requiring such rehabilitation is growing. Since the effects of rehabilitation show only if it is continuous, it is hard to keep motivation in monotonous rehabilitation environments. Therefore, with the use of latest virtual reality devices, it is...

Recently, flying robots such as multicopters are being used for various indoor and outdoor situations. The methods for operating a multicopter are either programming a flight path or using a controller. However, both methods require user preparation in order to control a multicopter as expected. This research proposes a finger-pointing multicopter...

The consultation algorithm for computer shogidetermines each moves by voting of multiple computer shogi players. In this research, we discuss some methods that optimize weights of player vote during a shogi game. To change the weight of player vote, we use the adoption times of each player vote and the evaluation value of the situation of the game....

We produced the eye-tracking system using the simple infrared ray web camera. Three basic movements were tested in order to use the eye-gaze input device. First, we watched four corners of the PC display and inspected whether a pupillary central coordinate accorded with the visual line direction. Second, we displayed three lines of sentences to the...

Tagging has emerged as a powerful mechanism that enables users to find and understand entities. However, there are three types of issues in traditional tagging systems. In this paper, we explore and seek a tag-algorithm that predicts tags of users and contents with a degree of relevance, which we called the tag ratio. We described our algorithm and...

The validity of the consultation algorithm is shown in computer shogi. The consultation algorithm determines a move by votes of computer shogi players. In this research, we discuss a method that changes player vote weights during a game. To change player vote weights, we use the selection history of the move and the evaluation function value. Exper...

In this study, we tackled the reduction of computational complexity by pruning the igo game tree using the potential model based on the knowledge expression of igo. The potential model considers go stones as potentials. Specific potential distributions on the go board result from each arrangement of the stones on the go board. Pruning, using the po...

Monte-Carlo GO is a computer GO program that is sufficiently competent without the knowledge expressions of IGO. Although it is computationally intensive, the computational complexity can be reduced by properly pruning the IGO game tree. In this study, we achieved this by using a potential model based on the knowledge expressions of IGO. The potent...

In this study, we tackled the reduction of computational complexity by pruning the igo game tree using the potential model based on the knowledge expression of igo. The potential model considers go stones as potentials. Specific potential distributions on the go board result from each arrangement of the stones on the go board. Pruning using the pot...

In this paper we use parallel processing to combine value functions in order to speedup reinforcement learning. We propose an asynchronous method of periodically composing Q table of local learning clusters to form global Q table. In this research, two approaches are implemented. First is discontinuance learning. Second is combination of value func...

The purpose of this paper is to propose an extended immune optimization algorithm using division as well as integration processing based on immune cell-cooperation and to investigate its validity by computer simulations. In the biological immune system, the immune cell-cooperation is a framework including MHC and immune network, the function of whi...

This paper describes a Competitive Coevolutionary Algorithm with potential model to solve Tsumego. The following three procedures achieve an effective evolutionary process. 1- The Go stones used in Tsumego are assumed to have a potential. 2- In any specific arrangement of Go stones on the board in Tsumego, the potential distribution of each interse...

In this paper, efficiency improvement of reinforcement learning using parallel processing for combination value function. We propose the method of periodically composing Q table of local learning clusters to global Q table. We apply this method to two applications. One is maze problem and an another is behavior rule detection problem for modular ty...

In this paper, We discussed about a controlling robot using iPhone(iPod touch). Recently, smart phones are popular that have a touch panel interface and various sensors. Such movement is the cause of IRT(Information and Robotics technology) progression. Smart control method is required for controlling robot by using such rich sensor device. Three a...

Congestion in parking lots is a common problem. While the behavior of the vehicles is known to have an influence on the congestion of the parking lot, this aspect has not been well studied. In this research we developed a traffic simulator to observe and study the influence of the behavior of individual vehicles on the overall congestion of the par...

Recently, it has become common for new words to be created and be used in conjunction with well known words to give new meanings on the Internet. While it is possible to guess the meaning of the words from the context, it is quite difficult to explain the meaning. In this research, using data from search engines, we focus on the frequency of the oc...

In this paper we present an approach to transform individual behaviour of homogenous sub-systems to yield desired global behaviour of the overall system. As a test bed we consider the brittle star robot as the system which is composed of homogenous modules (sub-systems). Using genetic algorithm, the rotational motion of the individual modules is tr...

In this paper we present a neural network controller for the modularbrittle star typed robot. Each leg of the robot is modeled as a fully connected neural network. The modules that make up the leg represent neurons, which are interconnected via synaptic weights. Motion is achieved by propagating phase information systematically across all the modul...

The process of evolution by natural selection enables search for best solutions by adapting to the problem domain as time passes by. While we have utilized GA to find suitable parameters for our control models it is worth mentioning that from a global viewpoint the models in turn have evolved from trial and error means of motion control, to CA-base...

This paper documents our ongoing efforts in devising efficient strategies in motion control of the brittle star-typed robot. As part of the control framework, each robotic leg consisting of series of homogenous modules is modeled as a neural network. The modules representative of neurons are interconnected via synaptic weights. The principle operat...

In this paper we report preliminary findings of using cellular automata (CA) as an underlying architecture in controlling the motion of a five-legged brittle star typed robot. Three control models were incrementally designed making use of genetic algorithm (GA) as well as co-evolutionary algorithm in finding appropriate rules for automaton. Simulat...

Recently, the research of leg type robot such as humanoid or dog type robot is becoming prosperous. However, there are no robots in existence which have the functionality to walk while keeping their balance in any kind of environment. In this paper we focus on the development of a five legged flexible myriapod robot in the form of a brittle star, a...

In a modular robot the global behavior of the robot emergence from combined action of all the modules. However in most cases it is difficult predict the global behavior of the robot, as it is not simply a summation of individual module behavior. In this research we model a five-legged brittle star type robot and incorporate forward locomotion in it...

Recently, a number of parallelized optimization algorithms have been proposed. We have proposed a co-evolutionary immune algorithm (IA) to solve the division-of-labor problems, in particular the n-th agents travelling salesman problem (n-TSP). In this article, we extend the co-evolutionary IA for a large-scale n-TSP with (1) an improvement for the...

In this paper, we introduce and describe the computational environment, which we have developed in this study, for cellular automata (CA). CA is a powerful method to understand and simulate the behavior of complex systems such as traffic jams, fluid crosscurrents, and natural disasters. In CA method, modeling of such a system or a phenomenon is to...

We have developed an efficient system (Multi-Auto-Transformation [MAT] vectors) for the removal of marker genes and to increase the regeneration frequency of transgenic crops without using antibiotic selection, reducing their possible environmental impact. The MAT vector system is designed to use the oncogenes (ipt, iaaM/H, rol) of Agrobacterium, w...

Recently, there has been significant research interest in homogeneous modular robots that can transform (i.e. reconfigure their overall shape). However, many of the proposed transformation mechanisms are too expensive and complex to be practical. The transformation process is also normally slow, and therefore the mechanisms are not suitable for sit...

This paper describes the compiler that we have developed in this study for high-speed simulation of cellular automata (CA). CA is a powerful method to understand and simulate the behavior of complex systems. CA is temporal and spatial discrete system. Updating many cells' state valves at high speed is needed for a high precision simulation. SIMD in...

The purpose of this paper is to evaluate an immune optimization algorithm using a biological immune co-evolutionary phenomenon and cell-cooperation. The co-evolutionary model searches solutions through the interactions between two kinds of agents, on the agent is called immune agent, which optimizes the cost of its own work. The other is call antig...

This paper describes about an automatic design method for an intersection that often causes a traffic congestion. We use cellular automata (CAs) for a traffic simulation, because CAs are able to reproduce a complicated phenomenon like traffic flow to decide local cell interaction. In recent research, acquisition methods of CA rule have been studied...

The purposes of this paper are to propose and evaluate an immune optimization algorithm using a biological immune co-evolutionary phenomenon and cell-cooperation. The co-evolutionary models searches the solution through the interactions between two kinds of agents, one of the agents is called immune agent which optimizes the cost of its own work. T...

Cellular Automata (CA), which is a method for analyzing phenomena of complex systems, allows us to construct many kinds of simulators such as road traffic simulators. When we design a CA model, we need a CA simulator for verification of its result. However, at the present, we need to develop a software simulator since CA dedicated computers have no...

In current transformation systems, a selectable marker gene is co-delivered with the gene of interest to identify and separate rare transgenic cells from non-transgenic cells. Since, during transformation, only a few plant cells accept the integration of foreign DNA, most of the cells remain non-transgenic. Usually, conditional dominant genes, whic...

Many systems have been developed for the removal of a selection marker in order to generate marker-free transgenic plants. These systems consist of (1) a site-specific recombination system (Cre/lox) or a phage-attachment region (attP) to remove the selectable marker gene and (2) a transposable element system (Ac) or a co-transformation system to se...

Competitive co-evolution algorithm can adaptively acquire solutions of a problem. However, in a problem which does not have the optimal solution, it needs to decide a set of effective solutions as the best solution. In this paper, we propose a competitive co-evolution algorithm with a packaging solutions to solve the problem. Our algorithm has two...

The purposes of the paper are to propose and evaluate an immune optimization algorithm inspired by biological immune cell-cooperation, and this algorithm solves the division-of-labor problems in a multi-agent system (MAS). The proposed algorithm solves the problem through interactions between agents, and between agents and the environment. The inte...

Cellular Automata (CA) are ideally suited for parallel processing, and have been characterized as easy to parallelize. Therefore their simulation has prospects of speeding up using SIMD (Single Instruction stream, Multiple Data stream). Furthermore a low-cost CPU has been had SIMD such as MMX technology. MMX technology is a kind of SIMD what permit...

The AC-rich motif, Pal-box, is an important cis-acting element for gene expression involved in phenylpropanoid biosynthesis. A cDNA clone (Ntlim1) encoding a Pal-box binding protein was isolated by Southwestern screening. The deduced amino acid sequence is highly similar to the members of the LIM protein family that contain a zinc finger motif. Mor...

The objective is to propose an immune distributed competitive problem solver using MHC and an immune network and to verify its validity by means of computer simulations. Our algorithm solves the division-of-labor issues and problems for each agent work domain in a multi-agent system (MPS) by two immune functions. First, the Major Histocompatibility...

The objective of this paper is to propose an evolutionary optimization algorithm using MHC and immune network and to verify its validity by means of computer simulations. Our algorithm solves the division-of-labor issues and problems for each agent's work domain in a multi-agent system (MAS) by two immune functions. First, the major histocompatibil...

Adaptive problem solving techniques such as neural networks and genetic algorithms become so popular in the AI field. The biological immune system is one of the adaptive biological systems whose functions are to identify and to eliminate foreign materials. In this paper, we propose an adaptive optimization algorithm based on immune model with immun...

As the neural networks or genetic algorithms, adaptive algorithms become popular and these techniques are applied to many kinds of optimization problems. The immune system is one of the adaptive biological system whose functions are to identify and to eliminate foreign material. In this paper, we propose an optimization algorithm based on immune mo...

In the field of artificial intelligence, the development of a method for the acquisition of game-strategies is one of the important issues. We deal with the acquisition of a game-strategy by using a competitive co-evolution approach as a search method. The competitive co-evolution is a mechanism of interactive improvement of solutions. However, in...

One of the important issues in the field of artificial intelligence is to develop the method on the acquisition of game strategy. In this paper, our purpose is the acquisition of game-strategy by using competitive co-evolution approach as a search method. The competitive co-evolution is the mechanism of the interactive improvement. However, in prac...

The combinatorial optimization problem is important problem for practical use in any factory. However the problem includes very difficult problem to solve automatically. One of the famous difficult problem of combinatorial problem is the Traveling Salesman Problem (TSP). This paper approaches to TSP using new information processing method that is a...

In this paper, we discuss about a competitive co-evolution model for acquiring strategies of games autonomously. To win a game, strategies of the game in various states are required. In the field of artificial life, genetic approaches are well known as an adaptive search algorithms. The genetic algorithm requires setting a fitness function and eval...

A new approach of finding the path to a goal automatically is proposed in this paper. For this problem, our approach uses the vibrating potential method (VPM) to provide the strategy for finding the path. The VPM is a new problem solving system we proposing, and has sufficient unity of fundamental mathematical expression to describe the movement of...

We have been proposing vibrating potential method (VPM) as a new mathematical model for an intelligent robotic system that can solve various engineering problems. In the VPM, an information field as energy field is defined and multi-agents can observe and communicate each other through the field. The agents can move on the field by getting the ener...

The paper presents an approach to terrain navigation by the vibrating potential method (VPM). In our VPM, the energy interactions on the problem considered are represented as the dynamical interactions of autonomous agents, which are defined as wave motions. We define the terrain information and vehicles as vibrating potential fields (VPF). The veh...

Univ.of the Ryukyus Univ.of the Ryukyus Tsume-go is life and death problem in Igo-game. It is considered the most difficult board-game for AI. In this paper, we introduce a competitive co-evolution algorithm with a direction of evolutional processes to solve the problem of Tsume-go game. We achieve an effective evolution process by three following...

The purposes of this paper are to propose and evaluate an immune opti- mization algorithm using biological immune co-evolutionary phenomenon and cell- cooperation. The co-evolutionary models searches the solution through the interac- tions between two kinds of agents, one of the agent is called immune agent which optimize the cost of its own work....

The purposes of this paper are to propose and evaluate an immune optimization algorithm inspired by a biological immune cell-cooperation and co- evolutionary phenomenon. And this algorithm solves the division-of-labor problems in multi-agent system (MAS) through the interactions based on such immune functions. The co-evolutionary models searches th...