Research Project Supported by Competitive Research Founding 2007, Annual Review 2007, The University of Aizu

Annual Review 2007 > Research Project Supported by Competitive Research Founding 2007

Research Project Supported by Competitive Research Funding 2009

General Research

Principal Investigator: Kazuyoshi Mori
Title: Parametrization of Stabilizing Controllers with Precompensator
Amount: \794,000

So far, we developed the parametrization of all strictly causal stabilizing controllers. The key idea was to use a precompensator of a stabilizing controller. We consider that this idea can be applied other parametrization of stabilizing controllers such as all stabilizing controllers that includes an integrator.
As a result of this project, we developed a new parametrization method of stabilizing controllers using precompensator. By using this method, we can obtain, for example, (1) all stabilizing controllers such that their relative degrees are a fixed number or more; (2) all stabilizing controllers such that they must have some fixed number or more of delay operators; (3) all stabilizing controllers such that they must have some delay operators of directions.
Principal Investigator: Nikolay N. Mirenkov
Title: Semantic surfaces for ubiquitous computing
Amount: \846,000

The project idea is to attach spatial connotation to surfaces that surround us and thus to create semantic surfaces for direct ambient interactions. Surfaces around us such as indoor walls, windows, product packing, newspapers and many others could be digitally enhanced with embedded spatial information for surface based interactions. Recognition of surface semantic could then be implemented through specialized optical readers that extract the spatial digital content and employ it for ambient interface control. To obtain the goal of the project we tried to answer to three main questions:
Systematic analysis of existing codes which can be used for semantic surfaces of different scales has been performed and the use of possible wearable devices with an embedded camera for ambient interactions has been considered. Some experiments with spatial surface connotation for enhancing multiple views of corresponding pieces of data/knowledge have also been done and preliminary answers to the above mentioned questions have been presented at three International conferences.
Principal Investigator: Jung-pil Shin
Title: A Study on pen Computing
Amount: \1,059,000

The purpose of our project, ”A Study on Pen Computing,” is the development of Pen Computing algorithm based on the Handwriting with high performance.
We have developed the algorithms ad systems for the following goal.
Principal Investigator: Incheon Paik
Title: Transformation of Non-Functional Properties for Automatic Web Service Composition
Amount: \1,015,000

Functional properties are discovered at the discovery stage and nonfunctional properties (NFP) are used mainly in the selection stage for service composition. Abstractness of NFPs has to be solved to be used in the selection stage for seamless composition. In this research, three levels of abstractness of NFPs (abstract, intermediate, and concrete) and transformation from intermediate level to concrete level were suggested. 　And using word similarity, translation from abstract level to intermediate level has been accomplished for complete transformation of NFPs. To solve vagueness of abstract NFPs, I adapt approaches based on not only ontology but also term similarity. Structure of NFPs is based on the definition of NFPs abstractness level, which has non-terminal compound terms. To evaluate effectiveness of term similarity metrics, vector-based and large corpus based approaches are investigated. Transformation performance based on precision over our test data set and ontology is evaluated.
Principal Investigator: Victor I. Ryzhii
Title: Graphene-New material for century Electronics:Theoretical and Numerical Studies of Novel Graphene Devices(II)
Amount: \1,162,000

The project was focused on the development of new basis for communication and computer systems of next generations utilizing carbon-based graphene, graphene bilayer, graphene nanoribbons, and multiple-layer graphene heterostructures in digital and analog circuits, terahertz and infrared wireless communications, remote sensing, and in numerous other devices. Such heterostructures will substantially surpass those on silicon and A3B5 semiconductor materials in speed and functionality.
The project supplements the activity in the framework of the project on grapheneon- silicon devices (with Tohoku and Hokkaido Universities) under the auspices of the Japan Science and Technology Agency, CREST.
The results obtained in 2009 fiscal year include:
Principal Investigator: Daming Wei
Title: EGC and EEG-based Stress Evaluation Utilizing Our Original Sensor Technique
Amount: \1,336,000

Experimental studies were conducted by driving simulation and long-term real car driving. Analysis were performed, resulted in publications as follows:
Principal Investigator: Maxim V. Ryzhii
Title: Computer simulations of electrical activity of heart and brain
Amount: \984,000

The aim of the project was modification and improvement of available software and development of original software on the base of the biological and computer models of the heart and brain, as well as building a high-performance computing system (workstation) for high-scale parallel numerical simulations and visualization of electrical disturbances and phenomena of human heart and brain. Simulations for solution of FitzHugh-Nagumo equations for excitable media, spreading of electrical excitation in Purkinje fiber on 3D slab structure, and excitation of spiral waves on 2D cardiac tissue sheet were performed. Results were presented at conference. Computer models of failing heart with precise physiological and structural parameters are under development.
Principal Investigator: Irina I. Khmyrova
Title: Analytical study and IsSpice simulation of heterostructure multi-gate devices
Amount: \835,000

To achieve the goals of the project analytical and equivalent circuit models were developed for multi-gate HEMT-based device. Resonant frequencies of plasma oscillations in ideal multi-gate HEMT have been evaluated. Performance of the multi-cantilever resonant sensor has been simulated using IsSpice software.
Principal Investigator: Konstantin Markov
Title: Establishment of Speech Recognition Technology and Development Program at the University of Aizu
Amount: \1,297,000

To have well developed Automatic Speech Recognition R&D program is the first step of establishing comprehensive Human Language Technology (HLT) research plan which will allow the University of Aizu to stay competitive, expand its research and education area, and increase its popularity and outreach. ASR is a multidisciplinary scientific field joining together speech and language science (linguistics, phonetics, etc.) with computer science and engineering (artificial intelligence, machine learning, signal processing, etc.). Studies in all these scientific fields are being conducted in various laboratories at the UoA and it is natural to take one step further and establish a joint, collaborative research program with the goal of advancing the spoken language technology and promoting new business opportunities and services for the local community.
Principal Investigator: Yong Liu
Title: ew Discoveries and their Applications in Evolutionary Programming
Amount: \906,000

This research is to establish a systematic design tool for photonic crystal design based on fast evolutionary programming. Photonic crystals are composed of periodic high and low dielectric constant nanostructures. They are beginning to find commercial applications, such as photonic crystal fibers possessing enhanced properties over normal optical fibers.
Photonic band gaps are essential to many real-world applications. To achieve a certain band structure with a required band gap is often a complex task of optimizing the lattice geometry, dielectric constants contrast, unit cell size, and filling fraction. Such a task is often too large to prevent the human designers from testing all designs. The optimal photonic band structure design is equivalent to finding the highest point on the design space. There are several characteristics on such a design space, which make evolutionary algorithms such as fast evolutionary programming better candidates for searching the space than other local optimal search algorithms.
This research will concentrate on three key issues in design of photonic band structures. The first issue is to develop a modified plane wave expansion for photonic band structures with arbitrarily shaped dielectric functions, and to simulate onedimensional and two-dimensional photonic band structures with pre-defined unit cell size, filling fraction, and arbitrary unit cell shape. The second issue is to develop an evolutionary system for optimizing photonic band structures based on fast evolutionary programming, in which two encoding schemes and the ordered mutations with constraints will be investigated. The third issue is to complete the application of the evolved photonic band structures on photonic crystal fiber.
Principal Investigator: Ian L. Wilson
Title: Neural Network Based Real-time Pronunciation Evaluation System
Amount: \1,024,000

The purpose of this research project was to develop the use of self-organizing map (SOM)-type neural networks in new pronunciation analysis software to be used in the Center for Language Research language classes. Neural network-based pronunciation feedback was integrated with feedback from ultrasound, MRI, and video, to be displayed in an interactive graphical user interface (GUI). The GUI was designed to run on iMac computers and uses students’ audio input from a headset microphone, as well as video input from the built-in camera on the iMac.
Principal Investigator: Taro Suzuki
Title: XPEF(XML-Based Planetary Exploration data Format) design and implementation
Amount: \1,136,000

The aim of the project is to design a XML-based data format called XPEF for the data archives in lunar and planetary science fields. It is proposed as a substitute for PDS (Planetary Data System) data format, a de facto standard of data format for lunar and planetary data but now outdated and hard to use, in order to provide flexible and easy-to-use data format for researchers in those fields.

We proposed the design of XPEF as a compressed data archive of data files, usually binary files such as image data, and XML documents that consists of a header for each data file and a file describing relationship between the files in an archive. We designed a XML schema for header information of data files by extracting necessary information from lunar surface image data taken by Clementine in 1994. We also implemented a conversion software of PDS files for image data by Clementine into XML header documents and image data.
Principal Investigator: Michael Cohen
Title: (UoA)^2:FiW;(virtual[world)music]
Amount: \976,000

University of Aizu and University of Alberta: Folkways in Wonderland; Virtual World, World Music
As immersive virtual environments and online music networks become increasingly popular, it behooves researchers to explore their convergence: groupware music browsers populated by figurative avatars. Collaborative virtual environments (CVEs) offer immersive experiential network interfaces to online worlds and media. We developed a virtual environment, based upon and similar to the ”Music in Wonderland” proof-of-concept by Sun Microsystems, that enables a place where avatar-represented users can go to browse musical databases. Music is a medium that is used for a wide range of purposes in different situations in very different ways in real life. Various systems and interfaces exist for the broad range of needs in music consumption. Locating a particular recording is well supported by traditional search interfaces via metadata, but improving search techniques via new strategies is a growing need.
Our music browser is a cylinder on which a rectangular map of the world is texturemapped, and tracks are placed according to origins of the respective songs, enabling location-aware browsing. The cylinder can be easily defined and projected in a 3D environment either declaratively or procedurally. Since Wonderland does not permit currently recumbency, the avatar always stands vertically. A sphere limits angle of view as subjects approach the poles; but on the other hand a cylinder can be effectively used to indicate geographical location using a suitable projection. An avatar can enter the cylinder, and click to listen track samples. Selected tracks are highlighted and multiple tracks can be heard at the same time when a user moves near from one track to near another, and in between sound may overlap. The system is collaborative: multiple users can hear the same music together, and they can hear each others’ speech via voice chat.
Keywords: Collaborative virtual environment, database browsing, groupware,Wonderland, spatial music, XML parsing, immersive environment
Principal Investigator: Robert H. Fujii
Title: Biologically Inspired Optimization Applied to Neural Motor Learning, Algorithm Engineering and Artificial Wind Tunnel Experiments
Amount: \1,143,000

Research Areas:
Principal Investigator: Tsuneo Tsukahara
Title: A Multi-band Wireless Receiver Using a Spectrum-Analyzing Method Suitable for Repester Nodes of Sensor Networks
Amount: \935,000

In this project, we proposed a multi-band wireless receiver using a spectrumanalyzing method suitable for repeater nodes of sensor networks. As the final goal after 3-to-5 year research, we would like to establish a“reconfigurable wireless communicator, ” whose frequency band can be changed according to communication conditions and/or a country ’s regulations using reconfigurable RF and baseband processors and downloadable software. This is a kind of cognitive radios based on SDR (Software-Defined Radios).
As the first step, in 2009, we concentrated on a receiver circuit design using a lowdistortion mixer and a spectrum-analyzing method. These circuits are the key to the development of a low-distortion reconfigurable receiver. Features we obtained are as follows:
These circuits are the key to the development of a low-distortion reconfigurable receiver. In the next step, we can proceed to a transceiver-circuit design including a transmitter relating to the 2010 application.
Principal Investigator:Emiko Kaneko
Title: Symposia on Global Citizenship
Amount: \1,021,000

This research was an attempt to provide students with an appropriate role model and ultimately enhance their motivation to learn English. For this purpose, we held five“ Symposia on Global Citizenship, ”inviting people from workforce and academia to give speeches in English about what was required for scientists and engineers to be successful in society. It was believed that listening to their speeches in English would motivate students 1) to achieve higher proficiency in English, 2) to consider what their career plan would be, 3) to make their campus life more efficient and fruitful, and 4) to make well-thought-out decisions when necessary.
For the purpose of measuring the changes in their motivation, a revised version of Gardener ’s Attitude/Motivation Test Battery (AMTB) was conducted before and after the five symposia. The comparison between pre and post test indicates that the more they realize the usefulness and importance of English, the more efforts they tended to make after attending more than 3 symposia. This is an encouraging change found among the core participants.
Principal Investigator: Shuxue Ding
Title: Blind source separation of non-sparse sources and solving communication problems using sparse representation
Amount: \1,085,000

Sparse representation of signals is a most general transformation in the sense that it is a shortest linear representation for the signal. This representation is based on the sparsity of signal, which says that most samples of the signal is zero or near zero valued. This property is of statistics, and the most pure video or pure audio signals hold such a property. That is, one can introduce a dictionary for representation of the signals, which includes a group of basic ”bases” or ”words”, i.e., ”atoms”. By a representation for a group of signals, one can choose several words from it and can use their combination to represent any signal. A very attractive feature is that, if a sparse signal is corrupted with noise or mixed with other signals, it will lose their sparsity or become less sparse. That is, to represent the signal one need more words from the dictionary. The sparse representation is just find a signal processing that can recover the pure signal by requiring the output with a higher sparsity.
In the researches of this year, we have focused on the researches of applying the sparse representation to the blind source separation (BSS) and the binary multiuser detection, beamforming and direction of arrival (DoA) estimation. We have further applied the developed method to blind spectral unmixing that is based on sparse nonnegative matrix factorization.
Principal Investigator: Hirokuni Kurokawa
Title: Skill Evaluation Support for Nuclear Power Plant Operator Training
Amount: \733,000

The goal of this research is to automatically and steadily track the trainees (nuclear power plant operators) in the training room, from multiple video sequences taken by a recording system. Our idea is using the combination of 3D particle filter and background subtraction where the particles are scattered inside the space of the training room and the likelihood of each particle is calculated based on histogram distan. The first version of the proposed tracking system has been completed.
Principal Investigator: Naohito Nakasato
Title: A compiler for Many-Core Accelerators and HPC in QP operations
Amount: \977,000

In this project, we propose a unified way of programming many-core accelerators. Our compiler system works like intelligent frontend software for each many-core accelerator. A final goal of this project is that with one input source code, we obtain optimized code running on variety of many-core accelerators. Along with the investigation of the new compiler, we investigate high performance computing in quadruple-precision (QP) FP operations. We already have many numerical problems that require QP or even higher precision operations. Due to its high computational costs, even a simple test of practical usage of QP has not been fully investigated. With high performance of many-core accelerators, we can test practical usages and implement a real application in QP. To advance the concept furthermore, we also consider hardware development devoted to QP operations.
Principal Investigator: Tuan Anh Pham
Title: Development and performance analysis of optical access systems using multiwavelength OCDM techniques
Amount: \1,147,000

The goal of this project is to propose cost-effective and feasible multiple-access technologies and architectures for the next-generation optical access networks. We especially focus on the novel optical code-division multiplexing (OCDM), the multiplexing technique that allows each user to use full resources of the optical fiber communication link. The major contribution of the project is the proposal of channel modeling technique, noise suppression and performance enhancement methods for OCDM-based access networks.
Selected publications resulted from this project:
Ngoc T. Dang and Anh T. Pham. Performance Analysis of Incoherent Multiwavelength OCDM Systems under the Impact of Four Wave Mixing. OSA’s?Opt. Express, Vol. 18, Issue 10, pp. 9922–9933, Apr. 2010.
Ngoc T. Dang, Anh T. Pham and Zixue Cheng. Beat Noise Cancellation in 2-D Optical Code-Division Multiple-Access Systems using Optical Hard-limiter Array. IEICE Trans. on Communications, Special Section on Optical Access Technologies, Vol.E93-B, No.2, pp. 289–292, Feb. 2010.
Ngoc T. Dang, Anh T. Pham and Zixue Cheng. Performance Analysis of Spectral Amplitude Encoding OCDM Systems over Linear Dispersive Optical Channel. IEEE/OSA Journal of Optical Communications and Networking. Vol. 1, Iss. 6, pp. 521–529, Nov. 2009.
Principal Investigator: Yukihide Kohira
Title: Design Automation of Clock Tree Synthesis for General-Synchronous Framework
Amount: \922,000

Nowadays, increases of the size and power consumption of a clock distribution circuit have become serious issues in the conventional clock synchronous framework in which the clock is inputted to registers simultaneously. On the other hand, the circuit performance is expected to be improved in general-synchronous framework (g-frame) in which the clock is not inputted simultaneously. In g-frame, the clock tree is synthesized so that the given clock schedule is satisfied. The circuit performance depends on the given clock scheduling method and the clock tree synthesis method. If these methods are not realized, the performance of obtained circuits is not improved or the obtained circuits cannot work correctly. In this project, the clock scheduling method and the clock tree synthesis method using the clustering method were investigated. In the clustering method, registers are classified into clusters. In the same cluster, it is assumed that the registers are inputted by clock at the same timing. Experiments show that the clock period in g-frame is improved compared with the conventional synchronous circuit and the size of clock tree obtained by the proposed clustering method is smaller.
Principal Investigator: Song Guo
Title: Adaptive Security Architectures, Algorithms, and Protocols for Wireless Sensor Networks
Amount: \985,000

Wireless sensor networks (WSNs) meet several operational challenges, such as energy efficiency in terms of maximizing the lifetime of sensor networks; scalability to a large number (thousands to millions) of nodes; survivability in certain environments where sensors are subject to compromise, capture, and manipulation by adversaries; support for dynamic addition/removal of sensors; and robustness to spontaneous interferences, collisions, and packet losses. Considering the network scale, the highly constrained system resource, and the fact that sensor networks are often deployed in unattended and hostile environments, we have developed efficient and scalable rekeying protocols with autonomous and adaptive capabilities for defending against node compromises in unattended sensor networks in this funded research project. The developed protocols have been also evaluated in terms of security level, resource efficiency and scalability by theoretical analysis and simulations. These results provide fundamental security services, covering key management, authentication, compromise detection, and revocation, which are essential for the successful deployment of sensor networks.
Principal Investigator: Kyoko Okudaira
Title: The University of Aizu Science Cafe
Amount: \322,000

University needs to find/attract excellent students for its growth. It is important to offer some opportunities for young students in order to raise awareness of sciencerelated subjects. However, parents have great influence on them at home.
Even if children show some interest in science or so, they do not know what to do next (who they can ask questions, what kind of study will be useful, etc.) when their parents do not have enough knowledge. It is a pity that parents cannot foster their children’s interest of science at home. Unfortunately, there are no scientific museums in Aizu.
Science Cafe can be one of the possible solutions to the problems described above. Although Science Cafe style has now become quite popular, we should not necessarily expect that only one or two activities leave an imprint on participants. Long-term, continuous activities will gradually bring fruits to our local community.
One of our duties is collaboration between junior high/high schools and the university for contribution to the local community. University can offer good opportunities for them and that is Science Cafe. At the cafe, you can ask questions freely in relaxed atmosphere with something to drink, listen to the speakers’ experiences in their childhoods, (This is what we only asked for speakers at the cafe) which is beneficial to both children and parents.
We applied for FY2009 “Development of Educational Science an dMath Program for Society” project, JST (Japan Science and Technology Agency), including Science Cafe as one of the activities. Our application was accepted and received good evaluation since students’ parents were also involved in our proposed activities for scientific education.
Principal Investigator: Hiroyuki Sagawa
Title: Microscopic Study of unstable nuclei
Amount: \989,000

Our goal is to investigate basic properties of unstable nuclei in microscopic models like Skyrme HF model and relativistic mean field theories. Especially the isospin structure of nuclear matter properties is very important subject to pursue in relation with nuclear matter EOS and neutron matter EOS. The role of hyperon is also important to understand in the field of strange matter in neutron stars.
In order to understand the EOS which can be applied for the study of astrophysical problems like neutron stars and supernovae, the experimental data of exotic beam factories like RIKEN Nishina Center will be very important. The theoretical study of exotic nuclei have strong impact on many field of physics from very small (femtosystems) to very large (astrophysical objects). We pursued this field of physics by using microscopic quantum many-body theories
We developed methodology and computer programming to study unstable nuclei, especially theoretical study of the isospin dependence of the pairing interactions and its effect of masses and excited states. We also studied the role of pairing correlations and also hyperons in the evolution of neutron stars.
Principal Investigator: Mohamed Hamada
Title: Design, Implementation and Application of Learning Style Index
Amount: \689,000

Learning Style Index is a method to classify the learning preferences for learners. Learning preferences can then help learners to find their better way in the learning process. It also can help teachers to adopt suitable learning materials for efficient learning. This project is concerned on the study, implementation, and application of a web-based learning style index. We also show a case study on the integration of learning style index into an e-learning system.
Principal Investigator: Gennadiy Nikishkov
Title: Curvature estimates for anisotropic rolled-up nanostructures
Amount: \916,000

Closed-form solutions under ordinary and generalized plane strain conditions have been obtained for anisotropic multilayer structures composed of cubic crystals under influence of initial strains. Constitutive matrix for material with cubic anisotropy was transformed by rotation around one of principal axis. Then equilibrium equations for nanohinge were formulated using the rotated constitutive matrix. Solution of the equilibrium equations provides closed-form estimate for the curvature radius of multilayer nanostructures. The solutions for estimation of curvature radius and strain components was applied to a bilayer In0.2Ga0.8As/GaAs rolled-up nanostructure. Curvature radius and strain components determined by the obtained solution are in good agreement with the finite element analysis. Java applet producing the solution for multilayer anisotropic structures was developed and placed on the Web.

Journal paper:
Y.Nishidate and G.P.Nikishkov, Curvature estimate for multilayer rolled-up nanostructures with cubic crystal anisotropy under initial strains, Journal of Applied Physics, 2009, 105, 093536-1-6.
Principal Investigator: Hitoshi Oi
Title: Resource management in virtualized systems
Amount: \911,000

The objectives of our project are as follows:

UBIC Research

Principal Investigator: Toshiaki Miyazaki
Title: Die-hard sensor network for monitoring disaster-hit region
Amount: \2,400,000

In the case of disaster, quick and detail surveillance of the damaged area is very important to make a rescue plan to save human life. However, the detail surveillance is usually done by hands in actual rescue scenes because there are no effective aids. Our goal is to develop a wireless sensor network (WSN) system, so called ‘die-hard sensor network’, which can monitor the disaster-hit region automatically by only scattering many sensor nodes in the region. The WSN has the following advantages; (1) no individual setup action is required in the field, (2) whole WSN system is never down and continuously monitors the disaster-hit region even if some sensor nodes involved in the WSN die, and (3) it has so scalable that we can expand the monitoring area and improve the sensing quality by only deploying extra sensor nodes as many as needed. In addition to these features, this system has a good cost performance compared to other surveillance systems such as a remotecontrolled helicopter. Thus, it can be applied to many critical disaster-hit regions, and we believe that our system must contribute not only to the rescue processes, but also to protecting the rescuers from secondary disasters. As the third year of the project, we concentrated on enhancing or developing the following technologies to realize the target system. (a) Reliability of wireless communications, (b) Location estimation of the sensor nodes, (c) Terminal software, (d) Sensor node design, and (f) Function alternation mechanism.
Principal Investigator: Qiangfu Zhao
Title: Card holder Authentication Based on Image Morphing and Machine Learning
Amount: \1,621,000

Consciously or unconsciously, we are using many kinds of“cards”in our daily lives. Among them, credit card, cash card, citizen card, driving license, and so on, are as important as our wallets. In most of the cases, we can use them safely without any trouble. However, with the rapid development of tampering technology, these cards may not be as safe as we expected. To protect the card holder, we propose in this project to re-shape the bio-image of the card holder, save it in the card, and reconstruct the image when the card is used. This year, we proposed a linear combination based method for generating reference images and a way for reducing the key size. We are going to file a patent based on the new results.
Principal Investigator: Takanobu Watanabe
Title: IT Service Application Studies for vitalizing Aizu region
Amount: \925,000

IT service is useful for vitalizing local region. However, IT service has not utilized well due to the lack of an opportunity to gather and share information about IT service. To solve above mentioned problems, we investigated latest IT service from three aspects: education, research, and business, and then give an opportunity to share those. To share the result of investigation, we published a booklet and developed contents of UoA undergraduate class; “ Basic Knowledge Course on Starting up Ventures ”taught by Prof. Kansen and held a regular meeting about with members belonging to local companies.
Principal Investigator: Wenxi Chen
Title: Development of an Innovative Sleep Monitor Using a Piezoelectric Sensor
Amount: \1,242,000

This study aims to develop a more sensitive and maintenance free sensor device based on a piezoelectric cell to monitor heart rate (HR), respiration rate (RR) and body movement (BM) during sleep, and to investigate the optimal position, such as head, back, hip and calf, for these vital signs detection.

Mid-Term Plan Research

Principal Investigator: Ryuichi Oka
Title: Free-viewpoint 3D TV
Amount: \1,064,000

We addressed a problem for realizing a system of free-viewpoint 3D-TV working in video rate speed of 30 frames per second and high vision size of pixels. The present state of the arts of 3D TV is described by three categories of technology. The first one is the classical method realizing by image projection on one screen by the output images of stereo cameras, and viewers watch the screen by through two filters each of them is corresponding to each eye. The second one is so-called a viewer-based 3D image reconstruction method. The most famous method of this category is the system based on so-called Ray-Space method proposed by Prof. Toshiaki Fujii of Nagoya University. This method does not construct 3D model so that this is called a viewer-based method. Users watch an image showing 3D world on TV. Moreover this method needs two or three hundred cameras which locate around an object. This system is still at the stage of research. Our proposed method belongs to the third category called 3D model-based reconstruction method. The core of system is realized by our method called two-dimensional Continuous Dynamic Programming. Our system seems promising because it needs only three cameras.
Principal Investigator: Takafumi Hayashi
Title: Cloud Service Integration using Content-Aware Networking
Amount: \1,996,000

The project studied network-centric cloud service integration using content-aware networking. The proposed solutions used a content -aware networking so called “ messaging networking, ” which is constructed as a structured overlay network. The proposed approach demonstrated a scheme to reduce the complexity and the maintenance cost of the system. The project shows the results not only for cloud integration. The results of the project present a secure scheme for a Smart Grid, inter-overlay integration for coming Post-Internet networking infrastructures. In the project, several network-centric service/contents integration are realized by using message mediation schemes and technologies. Various new scheme and new concepts are developed and published in this project as follows:
Several international conference papers were published and some sample implementation of the proposed scheme was constructed for the performance evaluation.
Principal Investigator: Vitaly Klyuev
Title: Using query expansion for query-oriented document summarization
Amount: \604,000

Search for scientific information on the Internet is not easy to perform utilizing the current search engines. The language to express queries is artificial: a set of key words. Its semantics is very far from the semantics of any natural language. Small snippets of texts presented by search engines in response to the user query cannot play a role of summaries of retrieved documents: In many cases, they are not indicative and users cannot judge the relevancy of documents without downloading them.
The main idea of our proposal is to use semantic relations between words applying the following tools: the WordNet database (English and Japanese), WordNet SenseRelate All Words, Stanford parser (for English), Mecab (for Japanese). To generate summaries, we utilized the sentence selection technique. The main results of our research are as follows: Key mechanisms to create summaries for English and Japanese were proposed; undergraduate and graduate students were involved in the project; four papers were presented at the international scientific conferences sponsored by IEEE.
Principal Investigator: Kenichi Kuroda
Title: An FPGA/GPU-base Scalable computing System for Open Source Scientific applications
Amount: \2,008,000

Three themes are focused. They are 1. Precision Improvement in SPH Simulation, 2. Rapid Matrix on an FPGA Acceleration System, 3. Performance Improvement of 2D-CDP Computation. The final direction of these topics is acceleration of software processing by hardware systems.
The SPH( Smoothed Particle Hydro- dynamics ) method attracts broad attention for behavior analysis of fluid and structural materials. The error by the cancellation of significant digits, which is important in SPH simulations, is improved by the offset value in the flow of simulation. The reduction of errors with grid partitioning and the NNPS( Nearest Neighboring Particle Searching ) method were studied. The proposed method improved the precision, and the cancellation of significant digits by collision was saved by the proposed method.
RapidMatriX is an accelerator of matrix operations. We implement it on an FPGA system. We proposed task scheduling and architecture for efficient data inputoutput. As a result, the throughput enhancement becomes pronounced and the execution is accelerated. For example, the speed was improved 1.26 to 1.83 times faster.
Performance improvement of 2D-CDP Computation is quite important in image matching applications. A new approach for optimizing and accelerating the computation of two-dimensional continuous dynamic programming (2DCDP) has been proposed. It is optimized by memory reduction, parallelization using OpenMP, buffer resizing, utilization of toggle-type buffers, and a realignment of buffering addresses for SIMD on multi-cores/multi-processors. The experimental results show that the computational time and the memory usage have reduced to about 1/4 and 1/5 of the original ones, respectively. Moreover, the concurrency of 2DCDP hot-spot is improved from 5.8 to 7.1 on a quad-core CPU with 8 threads.
Principal Investigator: Ben A. Abderazek
Title: In-Body Wireless Sensor Network System for Elderly Monitoring
Amount: \683,000

Recent technological advances in wireless networking, microelectronics and the Internet allow us to fundamentally modernize and change the way health care services are deployed.
Thus, the anticipated change and new services are urgently needed to help cope with the imminent crisis in the health care systems caused by current demographic, social, and economic trends in Japan and in most developed countries.
The ultimate goal of this project is to research and prototype an in-body wireless sensor network system to assist old people by monitoring their health status remotely, and to promote well-being by introducing smart in-body sensor based technology that allow medical professionals to initiate interventions in the home environment.
The novel system is based on a wireless micro-nodes implanted inside the body, where a node incorporate miniaturized sensors (physiological sensor like ECG, EEG, SpO2, etc.) and actuators.
Principal Investigator: Haruo Terasaka
Title: Research on Micro-scale Regional Climate Model
Amount: \1,493,000

This project aims to develop a high-resolution natural environmental simulation model and a forecast system. The results of the project are as follows.
Principal Investigator: Subhash Bhalla
Title: Interfacing Semi-skilled users with Spatial Web Services for Maps and General Applications for Medical Informatics
Amount: \958,000

Web is an interactive medium for information exchange. In recent times many web-based applications exist in different domains which require the use of a query language interface. Personal health records (PHRs) are being commonly adopted in health information systems. Patient centeric portals, such as Google Health and Microsoft health vault are becoming popular. Ability to query over this data plays an important role for users. Similarly, agriculture knowledge repositories and bio-medical information repositories have become available. The users need to query over large data. Many new applications also have geographic and spatial data components. There is a need to make spatial data query tools for semi-skilled users.
On the support side, query languages exist for database management systems. These languages require knowledge of programming skills. The proposed research aims at developing easy to use higher-level query interfaces for large data repositories. This effort will increase the ability of users to query information for diverse data resources.
Principal Investigator: Hirohide Demura
Title: Development of Space-borne Optical Instruments Systems: Data Processing/ Analysis and Hardware Design
Amount: \2,386,000

Web is an interactive medium for information exchange. In recent times many web-based applications exist in different domains which require the use of a query language interface. Personal health records (PHRs) are being commonly adopted in health information systems. Patient centeric portals, such as Google Health and Microsoft health vault are becoming popular. Ability to query over this data plays an important role for users. Similarly, agriculture knowledge repositories and bio-medical information repositories have become available. The users need to query over large data. Many new applications also have geographic and spatial data components. There is a need to make spatial data query tools for semi-skilled users.
On the support side, query languages exist for database management systems. These languages require knowledge of programming skills. The proposed research aims at developing easy to use higher-level query interfaces for large data repositories. This effort will increase the ability of users to query information for diverse data resources.