| Computational Intelligence and Systems |
Intelligent Control Systems |
Our research interests are optimization and machine learning for designing artificial intelligent systems in various fields. In the optimization, we focus on intelligent methods based on biological evolution and swarm intelligence. They include evolutionary computation and particle swarm optimization. In the machine learning, we focus on deep learning and reinforcement learning, which have attracted much attention. In addition, we apply them to real problems, and resolve problems encountered in the application.
Research themes: Research on optimization and machine learning for designing artificial intelligent systems
Keywords: Artificial intelligence/Deep learning/Reinforcement learning/Evolutionary computation/Genetic algorithms
|
Research into control systems began with feedback control and progressed to robust control.
Recent research is focusing on intelligent control, which aims at professional-engineer level control of systems.
We work with intelligent control, intelligent control-related information processing, and their computer applications.
In particular, we conduct research into robust control theories for control systems that can provide the required level of performance without complete knowledge of the objective systems to be controlled, hybrid system theories for handling control systems including complex logic circuits in integrated systems, clarification of mechanisms in living organisms that can flexibly respond to environmental changes, and the application of those mechanisms to control systems.
Research themes: Research on system and control theory to design intelligent control systems
Keywords: Intelligent control/System analysis/Control system design/Biological system
|
| Interactive Intelligence |
Multimedia Data Engineering |
Our lab develops intelligent agents that can interact with humans. In researching the nature of interactions, our aim is to develop a robot or agent that can start to acquire words and actions in much the way a human infant does, by watching, listening and interacting. We probe the secrets of intelligence by developing and operating working robots and simulators.
We also conduct research into spoken dialog systems as one type of interface with intelligent agents. For example, we are now developing a spoken dialogue system that supports elderly person in maintaining a healthy lifestyle and a multi-party intelligent tutoring system.
Research themes: Research on learning model for understanding intelligence, cognitive model and dialogue model
Keywords: Machine learning/Interaction design/Cognitive developmental model/Spoken dialogue system/Social presence
|
Along with advances in computers and networks, we often find we are surrounded by a huge amount of multimedia data – images, music, and video data, as well as images captured by digital cameras, downloaded music, and videos captured on video cameras. However, it is not easy to quickly and accurately retrieve the data we require from such massive amounts of multimedia data, even with the latest computers. We are working to resolve this problem by, for example, researching: (1) techniques that allow us to retrieve multimedia data based on its content and impressions, (2) techniques that allow us to retrieve similar multimedia data quickly, and (3) techniques that allow us to retrieve impressive scenes from video data by using facial expression recognition.
Research themes: Research on efficient, intelligent, and human-oriented processing of multimedia data
Keywords: Data engineering/Data science/Multimedia/Affective engineering/Facial expression recognition
|
| Image Processing |
Visual Information |
Most people use a large amount of visual information in everyday life. Along with recent advancements in computer performance, the use of digital images is rapidly increasing not only in industry and medicine but also at home, as can be seen with digital TV and 3D vision. Our research covers a wide range of image-technology including (1) real-time image acquisition and processing, (2) object detection based on feature extraction and instrumentation from acquired images, and (3) reconstruction and visualization of medical and other multi-dimensional images. These techniques are widely applied in industry and medicine; for example, to enable autonomous driving of cargo vehicles, defect-and-contamination inspection of products, computer-aided diagnosis of 3-dimensional CT images, and image synthesis.
Research themes: Research on image instrumentation and recognition, and its application in medical and manufacturing industries
Keywords: Image instrumentation/Image recognition/Machine learning/Design methodology of embedded systems/Industrial IoT
|
Different images are generated with different purposes such as computer tomography (CT) to “see” lesions. It is difficult to understand all visual information and derive potential information from such visual information. This research is to use visual information to help experts obtain effective information by signal/image processing, pattern recognition, computer vision and deep learning. The following researches are ongoing: (1) Log features extraction for experts in studying wood strength, (2) Image inpainting of long-standing damaged artworks for artists, (3) Detection of lung nodules and classification of their benignancy or malignancy for radiologists, and (4) Features extraction from fundus images for ophthalmologists.
Research themes: Effective visual information extraction to experts of different fields
Keywords: Signal and image processing/Pattern recognition/Computer vision/Machine learning
|
| Computer Systems |
Software Engineering |
Nowadays computer technology is embedded in various equipment, and the expansion of computer applications makes them even more critical as indispensable tools. To enhance the performance of the computer and make it applicable to practical uses that may have previously been thought to be impossible, we research a wide range of fundamental technologies related to computer systems. For example, we have been designing 1) a new microprocessor, which is the central device of a computer, 2) a high-performance parallel/distributed computer system, which consists of multiple processors/computers connected with each other through a network, 3) a secure computer system with special facilities to maintain or improve security, and much more. We are continuously developing novel techniques for entire computer systems that include not only hardware but also such basic software as compilers and operating systems.
Research themes: Research and development of organization schemes of high-performance computer systems and their elemental technologies
Keywords: Computer architecture/Parallel processing/Microprocessor design/Storage system architecture/Multithreading architecture
|
With the expansion and increased complexity of software development, software development methodologies cannot keep up with trends in software development. Insufficient methodologies resulting from this cause various problems such as deadline delays, excessive development costs, and poor quality. We are extracting insights and methodologies for software quality improvement from past development histories using software repository mining techniques. We are researching a wide range of areas from analysis of open-source software to data analysis of software processes in collaboration with industry. Specifically, we research (1) techniques for detecting faults in software modules using text analysis and (2) techniques for the simulation of software development processes.
Research themes: Software defect prediction using machine learning technique
Keywords: Software repositories/Software defects/Source code analysis/Machine learning/Software testing/Clone code
|
| Information Security |
Communication Systems |
It is virtually impossible to get along in society today without our ubiquitous information and communications systems. To enable secure use of these systems, information security technology has now become essential. The technology is related to a wide range of fields from such basic mathematical theories as cryptology and coding theory, to network technology and signal processing technology for digital content. We research: (1) intrusion detection systems (IDS), spam mail countermeasure technology, and other network security countermeasures (2) copyright protection technology for digital content, such as digital watermark, and (3) personal authentication technology such as fingerprint authentication.
Research themes: Information security for privacy protection
Keywords: Network security/Secure protocol/Technology for copyright protection/Personal authentication
|
Imagining a social life without information and communication technology (ICT) is difficult today. Not only high communication efficiency but also high energy efficiency is demanded of future communication networks. To establish such highly efficient networks, we research (1) modulation and coding schemes to enhance communication reliability and increase resistance to signal fading and noisy channels, (2) medium access control protocols to reduce data packet collisions in contention-based networks, and (3) sleep control technology and network protocols to achieve high energy efficiency for data collection networks with massive numbers of wide area sensors.
Research themes: Research and development of communication systems based on sustainability
Keywords: Synchronization/Modulation and coding schemes/MIMO/Medium access control/Multiple access/Local positioning systems
|
| Distributed Systems |
Educational Technology |
A distributed system is one in which many computers are connected through a network. The internet is a huge and familiar example of a distributed system. Our laboratory researches operation and management techniques for distributed systems and e-learning. To maintain distributed systems in a secure and reliable state, among other operation and management techniques, we need security practices and design methods to make these systems less prone to failure. E-learning enables these techniques to be used for educational activities in environments where the computers of individuals are connected to a network. Distributed systems are expected to facilitate an ever widening range of educational activities in the immediate future.
Research themes: Research and development on the operation and management technology of network and computer infrastructure
Keywords: Internet operation and management technology/Distributed system management/Infrastructure visualization/e-Learning
|
Today, students use their own information devices, such as laptops, tablets, and smartphones for their learning activities. As a result, the role of educational computer systems in university is expanding from simply providing computer lab access to the support of various aspects of university life, including learning and research activities. In this lab, we currently focus on educational technologies: 1) digital learning environments based on distributed system architecture, 2) data collection and analytics in learning environments, and 3) information system infrastructure for educational systems. For example, we have developed learning environments with virtualization technology, automated lecture capture systems, and identity-and-access-management systems for user authentication.
Research themes: Research and development on educational technology
Keywords: Online learning environment/Learning analytics/IT infrastructure for educational environment
|
| Human Information Technology |
Cognitive Behavioral Science |
Our research target is the realm of the relationship between human and machines, namely, computers, and is known as Human-Computer Interaction. This field has been growing in importance as computers increase around us.
We are developing methods for utilizing complex information technology efficiently, comfortably, and to good purpose. Specially, we manipulate and analyze 1) technologies for recognizing and supporting human movement and actions based on the prediction and understanding of human tasks, 2) environments for simply and comfortably utilizing complex computer systems functionality, and, 3) technologies for making various human learning processes more effective.
Research themes: Research of supporting systems for human activities with information technologies
Keywords: Motion sensing/Machine learning/Agent design/AR and VR/IoT
|
With the development of information technology, the machine and systems in our lives are becoming increasingly complex. We research the creation of design guidelines for improved human-machine interaction by clarifying the human cognitive and psychological characteristics. The aim is to elucidate individual differences in human cognitive and psychological characteristics by experimental psychological approaches, and to provide an interaction design platform for equipment that will satisfy individual users with various characteristics. One of our applied research target is about driver-vehicle interaction, we examine the relationship between driving behavior and driver’s cognitive and psychological characteristics using a driving simulator.
Research themes: Research of cognitive information process on human-machine interaction
Keywords: Cognitive and psychological characteristics/Individual differences in human characteristics/Cognitive science and engineering
|
| Human Interface |
Data Science |
Human interfaces (HI) are the focus of our research. This includes any computer, robot or similar contrivance placed between one human and another or any such technological domain in which problems occur. As machines with excellent functions are worthless without good usability, HI is essential in machine use. Equipment with a good human interface is user-friendly, more efficient, and causes fewer errors. We seek to develop user-friendly HI devices that are highly relatable to the general public. Our goal is high-performance HI devices that enable both human-to-machine and human-via-machine communication.
Research themes: Research on user-friendly and comfortable interaction
Keywords: Human interface/Interaction design/Media communication/Mobile interaction/Awareness
|
Data science is the study of statistical and mathematical techniques for dealing with data and its applications. It enables us to reveal the truth of situations using data from perspectives other than those of traditional theories and methods. To date, IT, in the business field, has mainly managed numbers by aggregation. At present, the amount and types of data are exploding in volume. It is required that data be used effectively. This applies not only to corporations, but also to governments and educational institutions that seek to improve the quality of services for residents and students. In our laboratory, we analyze actual data such as web data and sensing data, and study technologies that contribute to actual business and social problems.
Research themes: Research on technologies to solve business and social problems using statistics, machine learning and programming
Keywords: Text data analysis/Interaction models/Biological information modeling
|
