Theoretical Information ScienceIshihara Laboratory
Constructive Mathematics
z z
AnalysisHilbert spaces, Banach spaces, Linear operators TopologyFormal Topology, Neighbourhood spaces Intuitionistic LogicProof theory, Model theory Constructive Set TheoryIZF, CZF, Type theory Constructive Reverse MathematicsContinuity, Compactness Computability TheoryComputable functions, Lambda Calculus Complexity TheoryPolynomial time computable functions
Mathematical Logic
z z z
Computation Theory
z z
Theoretical Information Science Algorithms Asano Lab. Computational geometry Given a number of points in the plane, partition the plane into regions by which point is closest. It is called Voronoi diagram. It also has a number of applications.
Research ResearchTopic Topic "Computational "ComputationalGeometry" Geometry" "Algorithms" "Algorithms" "Combinatorial "CombinatorialOptimization" Optimization" Algorithms Developing fast algorithms is of recent interest. Especially interested in designing algorithms using limited working space. Combinatorial Optimization Digital Halftoning: How to optimize photos for printing it using limited number of inks. This is a problem of combinatorial optimization.
Find every copy of a pattern in the left figure in the right figure. Compute it using only constant working space.
IntroductiontoUeharalab
[Keywords]
Algorithm,ComputationalComplexity, GraphTheory
GraphIsomorphism
[SubjectofStudy]
Algorithms solvingintractable problemsinsomeway Intractableproblems: NPhardproblem(e.g.,longest pathproblem) Unknown U k complexity l i (e.g., ( graphisomorphismproblem) Hugedata(e.g.,Webgraph, Bioinformatics) Geometricobject(e.g.,Origami)
Maekawa Deviland KawasakiRose
[email protected]
Restrictedgraphs Randomizedalgorithms Approximation Computationalcomplexity Random R d graphs h Geometricmodels
Tilingofanetthat canfoldtotwo incongruentboxes.
http://www.jaist.ac.jp/~uehara/
Hiraishi Laboratory
We study how to build models of complex real-world real world objects in a mathematical way.
Theory and Application of Formal Modeling
Target Areas
delay 2 sec.
Systems Biology
on/off
y 1
l0 x 1 y 1 y 10 x2
y 10
x : 0
l1 x 1 y 1 x2 x2 l2 x 1 y 2 y5
sensor
l3 x 1 y 2 x2
y5 x : 0
Hybrid Systems
Service Processes in Nursing & Caregiving
ResearchTopicsinOgataLab.
Mathematicalmodels,specificationandverificationfor distributedsystems,andtoolssupportingthemarestudies.
1. Mathematicalmodels (ObservationalTransition Systems,orOTSs) 3. Verification (Theoremproving, Modelchecking,)
X = Y Y = Z X = Z f(X) = g(Y) X = A Y = B f(A) = g(B)
2. Specification (CafeOBJ,Maude,)
ceq pc(want(S,I),J) = (if I = J then ws else pc(S,J) fi) if c-want(S,J) .
(pc(s,i) = cs pc(s,j) = cs) (i = j)
4. Tools (Translators,Provers,) Distributedsystems (Dist.mutualexcl.protocol, Communicationprotocol, Ecommerceprotocol,)
Theoretical Information Science
http://www.jaist.ac.jp/~omote
Information Security Group Omote Lab.
The network system is complicated in recent year and hence the number of terminals and the amount of traffic in the system become so huge. Thus, not only security of each terminal but also security of the whole system become important. Our laboratory studies security modeling of network systems, based on cryptography and probability / statistics theory.
Recent Topics
Sensor Network Security
Probabilistic key sharing, Self-healing, Secure aggregation
Countermeasure against Illegal Access
St h ti modeling, Stochastic d li Si Simulation l ti experiments i t
Cloud Security
Proof Of Retrievability POR, Data restoration, Data integrity
ComputationalNeuroscienceLaboratory PI:HirokazuTanaka,Ph.D.
Ourmission:Understandtheinformationprocessingalgorithmsandrepresentationsused inthebrainthroughcomputationalmodeling,biologicalsignalprocessing,and psychophysics. InformationProcessingof Computationalmodelingof Spikingneuralnetworks sensorimotorcorticalnetwork
Analysismethodsforbiological signalprocessing
Humanpsychophysicsbasedon computationalmodels
Human Information Processing
Acoustic Information Science / Akagi Laboratory BASIC CONCEPT
Speech production and perception are humans activities. Study knowledge on speech production and perception as humans activities and construct useful models for advanced sound processing systems.
RESEARCH TOPICS
Production
Investigating the relationship between speech spectrum and shape of vocal tract, synthesizing natural speech with non-linguistic information (e.g., individuality and emotion), and synthesizing singing voice
Perception
1
Recognize speech in real-world conditions, realizing cock-tail-party effect, to enhance speech, and constructing models of effective speech
Human Information Processing Acoustic Information Processing
Modeling of human auditory system Application for auditory signal processing Human auditor filterbank
Lab: http://www.ais.jaist.ac.jp/index-j.html Per. http://www.jaist.ac.jp/~unoki/
Unoki Lab.
Investigation of human frequency selectivity Estimates of auditory filter Masking experiments
Modeling of cochlear filtering
Denoising and dereverberation method Bone-conducted speech
Physiology Audio watermarking technique Information Science Psychoacoustics
YOSHITAKA Lab.Image Information Processing
Realizing Novel Framework on Video Processing that Fits Human Ways of Perception
Studies on image processing system including issues on human factor and human-computer interaction Pursue feature extraction / retrieval method based on human way of image/video recognition or understanding, focusing on emotional information Reorganization of read-world information space and Web information space for better accessibility with a uniformed way Research Topics Emotional Information Extraction based on Cinematography Video retrieval based on emotional information Visual data organization based on eye movement Automated Social Filtering based on Gaze Detection Speaker Tracking System using Tagged Microphone Web Information Reorganization based on Human Activity in the Real World, and more...
Robotics Laboratory Convergence of IT and Robotics
URL: http://www.jaist.ac.jp/robot/ Professor
Self-organizing Robot Swarms
Sensor Networks for Ambient Intelligence
Humanoid Service Robot
JAIST Active Robotic Walker
Human Information Processing Robotica Lab.
Asano Laboratory
Our laboratory studies robotics and control engineering, especially humanoids and their motion control and technologies. We aim to gain a deeper understanding of skillful and adaptive mechanisms of humans and animals through robotics research, and to develop novel machine systems that can achieve more advanced movement than living organisms.
System Control Theory
Nonlinear Hybrid Dynamical Systems
Limit cycle walking
Dynamics Based Control Biologically Inspired Technologies
Computational Engineering and Science / Matsuzawa Lab. Research Brief Flow can be found everywhere in air, water and etc. Computational Fluid Dynamics (CFD) is the main research topic in this laboratory. CFD aims to simulate and clarify flow by numerical method. Bio-fluid flow and High performance computing is the active research areas. Bio-Fluid Flow The blood flow impacts mechanically blood vessel wall. This mechanical stress is regarded as one of cause of circulatory disease e.g. aneurysm, dissection. Clarifying blood flow and mechanically blood wall through blood flow leads to a more advance medical support. Heart Simulation Circulatory Simulation such as Aneurysm and Dissection etc Blood Flow Simulation with Medical Devices Nasal Flow Simulation High Performance Computing and Visualization Super parallel computing are expected to accelerate CFD . The laboratory is investigating on performance improvement on parallel computers and the large scale data visualization technique. Development for high performance linear equation solver Scientific Visualization method optimized for flow
Flow of Aortic Aneurysm
Flow and Temperature distribution in nasal cavity
Cray XT5
Simulation Science / High Performance Computing (HPC)
Maezono Group
using Super Computers owned by JAIST.
Search by Google!
HPC applied to Computational Materials Science
Were using
Quantum Physics Information Science
World fastest Super ComputerK
Maezono group is a representative user as a specially selected large-scale application 2012.
Hand-made PC cluster ~ 800cores constructed and working now!
Simulation Science as a 21century Industrial Science. Experts required not only on Modeling Sci. but Comput. Sci. as well.
2009. 4.26
Shimazu-lab NaturalLanguageProcessingand
ApplyingInformationSciencetoSocialSystems
TechnologyforprocessinglanguagessuchasJapanese,EnglishandVietnamese NewApproachtoSocialSystemsbasedonIS,specificallyLegalEngineering Basic Technology
Analysis, Generation, Disambiguation, Corpus, Machine Learning Machine Translation Spoken Dialog Processing Info Dynamic Explanation Dialog Robot Web Text Processing Web Interface
Cut asparagus to length of 3 4 cm.
Generating Animation
ApplyingInformationScienceforDescribingandDesigningSocialSystems
Legal Engineering
Methodology to systematically help make or change laws and rules, to be consistent and highly complete Methodology to design law implementation information systems based on laws or rules Methodology to keep information systems and rules of organizations in compliance with laws and standards Analysis of Legal Documents Formal Description of Legal Texts Consistency Verification of Legal Documents Structural Rewriting of Legal Documents
Shirai Laboratory
Theme Natural Language Processing (NLP) (especially) corpus-based NLP
Acquisition Knowledge Utilization
NLP Application Corpus
(Huge collection of text)
Statistical Information
Research Topics Statistical natural language analysis Induction of new word senses Interactive question-answering system Animation Generation of cooking recipes
Logic and Language in Agent Communication Tojo Lab.
1101101010011001011100
Language
Logic
Agent communication: Logic of knowledge and belief, Communication channel Legal reasoning: Verification, Multiple if-then structure Language evolution: Cognitive bias, Language acquisition and change Grammar in music: Tree structure, Chord progression rules
Entertainment and Intelligence
Professor dr. Hiroyuki Iida
OM-search vs. Minimax search H.Iida (1993) vs. von Neumann (1928) Human vs. Computer Yonenaga vs. Bonkras (Jan 14, 2012)
Three Masters model [Iida 2000] Presents three aspects of games
Information dynamics vs. Information theory H.Iida (2011) vs. C.E.Shannon (1948)
Game-refinement theory vs. Game theory H.Iida (2004) vs. J. Nash (1948)
IKEDA lab
Research Unit for Computers and Games
Game AI, as our rival and teacher
Using Genetic Algorithm...
By Learning, Reasoning...
Build the research basis on good game AI
Developing Distance Learning Environments for Effective Computer-mediated Interaction
Research Management Implementation
Planning Research Activity Self-directed Learning Human Information Processing Distance Learning System ICT, Audio-Visual Educational/Learning Technology System Design
Tanaka Lab.Evolution of computer architecture by Hard/Soft cooperation
Processor & Memory Architecture
Highly
functionality & Reconfigurability Low energy by Hard/Soft cooperation Heterogeneous g instruction-set multicores
Embedded systems y
Development of real-time processor Trial manufacture of processor LSI Development of real real-time time OS Conforming to ITRON interfaces
LSI design environment
Object-oriented SoC design by Java High-level synthesis & HW libraries
LSI for Advanced Information Society
Prof. MINEO KANEKO, Assist. Prof. TSUYOSHI IWAGAKI Large Scale Integration (LSI) = Brain, nervous system, sensory organ of information equipment = Performance and Reliability of LSI support sound and sophisticated information society
Billions of transistor on a single chip to form System-on-Chip (SoC) Optimization problems of huge size Design methodologies, LSI architectures for designability Propagation delay and its variation affect and limit system performance Locations of data and operation determine performance Computational algorithm considering ``location'' and ``move'' Correct operation under static and dynamic delay variability LSI computation with robustness against ``variability Safety/Reliability of information society is supported by reliable LSIs Reliable LSI computing Fault tolerant LSI computing, high quality LSI testing
Shinoda Lab
Internet Research Lab.
Every Thing About Internet
Empirical verification of large network systems StarBED Ubiquitous System Simulator
R&D of core technologies for next generation network systems. (E.g. routing, P2P systems, )
Hackers with artistic taste wanted !!!
School of Information Science
LIM Laboratory (Ubiquitous Networking Research Group) Associate Professor Azman Osman Lim
Research Goal Home Network
Summer
Outside Outside
Summer Winter
Outside
37C
28C
Inside
20C
Air Con
Inside
0 C
Winter
Inside
Sensor
Sensor
Window Internet Sink Gateway Server
Intelligent Home expects a network of electronic devices, computers, appliances, sensors and actuators from different vendors are interoperable with each others to assist people live on their own comfortable. Wireless Sensor and Actuator Network (WSAN)
Sensors gather information about the state of physical world. Actuators react to this i f information ti by b performing appropriate actions. Thus, WSAN enables cyber systems to monitor and manipulate th b the behavior h i of f th the physical world.
Data aggregation is performed locally
In Home Network, the sensors pass information among themselves and to a sink. The gathered data is analyzed to infer activities of daily living, which can give important clues to a person's state of comfortable, relax, restful, and pleasant.
In-network Aggregation Technique
Application Framework Application Support Sublayer y Network Layer Applications Data Representations Aggregation Functions Routing Protocols PHY and MAC Layers Towards the Sink
In-network Aggregation is a
particular in-network data processing technique that exploits the characteristics of the wireless medium (especially broadcast communication channel) in order to increase the network capacity.
A single sensor failure can break the data flow
PHY and MAC Layers Towards the Sink
Data follows multihop path to the sink
Sink Sensor Actuator
Network Service & Network Experiment
Research Topics Server development Performance Evaluation Testbed building Experiment automation Keywords HTTP, sockets, threads Bandwidth, latency, jitter lex&yacc, make, svn
Chinen
socket bind listen accept fork close read write close
School of Information Science
Information Theory and Signal Processing Laboratory
Prof. Tadashi Matsumoto, Fellow, IEEE and Assistant Prof. Khoirul Anwar We have sought for the possibility that the BP algorithm for the decoding of turbo codes be used in eliminating severe ISI due to multipath propagation. This equalization concept, not relying on big brain, but connecting small brains over which BP is performed, is referred to as Turbo Equalization. Our mission is to make impossible mission possible! 2000 Time Domain Sliding Window Algorithm Complexity=O(L3): L=number of the path 2002 Extended to MIMO Complexity=O(R3L3): R=number of Rx antennas 2003 Time Domain Block-wise Processing 2005 Frequency Domain Algorithm Complexity=O(logN): N=Frame Length, Independent of the channel! 2006 Industry standardization meetings started feasibility investigations!
IEEE Japan Council Award of Meritorious Contribution to the Society IEEE Vehicular Technology Society James R. Evans Avant Garde Award Thuringen States Research Award on Advanced Applied Science Finnlnd Distinguished Profesorship 2008 IEICE Best Paper Award 2009 Telecom System Technology Award by the Telecommunications Advancement Foundation IEEE Fellow, 2010IEEE ComSoc Exempelary Reviewer Recognition For more detailes, please see: IEEE VTS Distinguished Lecturer www.jaist.ac.jp/is/labs/matsumoto-lab UK Royal Academy Visiting Fellow Award
Time
Information Theory and Signal Processing Laboratory
Associate Professor Brian Kurkoski
http://brian.kurkoski.org/
7 6
Pursuing the fundamental limits of communication systems
Channel Model transmitter receiver LDPC codes achieve capacity
5 4 3 2
C = max I (X ; Z )
pj
1 0 0
Shannon capacity
Error-correcting codes
Lattices codes
Convex/concave optimization
Put theory to practical use Practical problems motivate theory
Communications mobile data fiber optic
Storage Flash memory/SSDs Hard drives
Multimedia security digital watermarking image hashing
2D detection 2D bar codes
Fundamentals begins with Fun
Suzuki Lab.
Software Constuction and Comprehension Architecture and Component based SE
Software development becomes more and more complex by the following reasons:
increase of functional requirements shorter durations collaboration of heterogeneous applications on WEBs We need a new mechanism for more sophisticated reuse. Reuse of structures and behaviors are achieved by Components components for PCs for mobiles architecture
Replace of components allows us to add/change functions easily
We need highly reliable components and connection mechanisms for applications more useful and dependable. Construction supports to detect improper connections. Different vender may use different protocol for interactions.
eg: mismatch in data order
setX(x) setY(y) do()
do(x,y)
an adapter is generated automatically
Comprehension supports to build new components.
public class MenuBarFactory { private ... instances; public ... newInstance(...) { ... mb = new ...; ... m = new ...; mb.add(m) instances.add(mb); } class propertylist instances m methodlist newInstances mb instances
Graphical display allows us to detect which parts should be changed.
Computing Thoughts
Ogawa Laboratory
Hirokawa Laboratory: Theory of Computing
MISSION
investigate mechanism of computation analyze computational properties automatically
TOPICS
conuence, termination, complexity analysis, ... program
e.g. quick sort
COMPLEXITY TOOL
complexity
O(n2 )
KEYWORDS
term rewriting, functional programming, SAT/SMT student who likes functional programming is most welcome!
CafeOBJ, OCaml, Ruby, Scala, ...
Somemura Laboratory
mission Education research concerning the global leader training program in an ICT field
Background
Progress of the globalization and the knowledge based society
Graduate education which brings up the leader who can play an active part in international society is essential.
Purpose
Personnel training, which raises the talented people who can assert a standard of Japan and play an important role in technical committees, such as ISO, IEC, and ITU, for the international standardization of an ICT field.
Needed knowledge is broad culture, the communications skills in English, debate ability, mental faculties, and the deep insight to a special field of study.
Global leader training program in an ICT field
Study abroad
Goal
Internship
Collaboration among Universities
Establishment of the graduate education which brings up advanced people who can play an active part in an ICT field
Research Topics Green ICT toward low-carbon society
Methodology for environmental impact reduction effects by using ICTs Methodology for Biodiversity Environmental Accounting / CSR Accounting
