1. Home
  2.  > Programs & Courses
  3.  > Graduate School of Science and Engineering : Information and Computer Science (Laboratories : Intelligent Information Processing Laboratory)

Graduate School of Science and Engineering
Information and Computer Science

Intelligent Information Processing Laboratory

We are aiming to realize people-friendly intelligent robots


Hirokazu WATABE
Telephone : +81-774-65-6946
Office : KC-211
Database of Researchers
[Associate Professor]
Telephone : +81-774-65-6948
Office : KC-213
Database of Researchers

Research Topics

<1> Research on concept base automatic construction/refinement methods
Automatically extract various concepts from electronic dictionaries and web information, construct a concept base of over 200,000 words, and automatically refine it.

<2> Research methods to quantify the degree of association between concepts
From the correlation of meanings of words defined by the concept base, make it possible to quantify the relationship between concepts and associate and recall them.

<3> Methods to implement a common-sense judgment mechanism for time/location
Make natural conversations possible by implementing a common-sense judgment mechanism for seasons, time, and location.

<4> Methods to implement a common-sense judgment mechanism for emotions/senses
Make natural conversations possible by implementing a common-sense judgment mechanism that feels happiness, sadness, or the beauty of sunset.

<5> Research on intelligent conversation mechanism
Make natural conversations with humans, like greetings and small talk, possible while understanding the meaning of speech with intention understanding and common-sense judgment.

<6> Research on document classification/summary
Based on concept processing, compare many documents such as articles, classify them into categories, and create document summaries.

<7> Research on chart meaning comprehension
Make it possible to understand the meaning of information expressed as various types of charts in the same way as humans do.

<8> Research on intelligent search methods; web, etc.
By associations using concept bases and common-sense judgment, make information searches on the web, etc., more intelligent and easy to use.

<9> Research on signboard understanding
Extract, recognize, and understand character information, image information, and map information contained on signboards and understand the meaning of the signboard as a whole.

<10> Research on environment understanding based on maps and images
Understand simple overview maps, recognize and understand roads and buildings from the corresponding image data, and create an actual map.
<11> Research on intelligent robot
Implement intelligent robots that autonomously move, gesture, and change expressions based on spoken conversations with humans.

Errand robot Robovie!
Environment understanding and autonomous movement

Must have the ability to understand its environment and move autonomously
  • Recognize objects such as signs, buildings, people
  • Recognize the meaning of characters and diagrams in signboards, markers
  • Understand guide maps and comprehend routes
  • Compare guide maps to the actual environment and move autonomously

Research Contents

Research to make computers intelligent is important for the advancement of information technology in the 21st century. In this research laboratory, our goal is to implement intelligent computers that can appropriately pick up on people’s intentions and to advance research on intelligent mechanisms that are the basis of that goal. To implement intelligent computers, we must develop mechanisms that can make common-sense judgments like humans such as judging amounts (big, small, etc.), time, and locations, and emotional/sensory judgments such as emotions and the five senses; judgments that are logical relationships with language concepts. The association mechanism implemented by measuring the degree of association with the concept base plays a particularly important role. As an application for intelligent computers, an intelligent robot is being developed that can hold spoken conversations with humans, understand its environment from images, and move autonomously.


  • Knowledge processing
  • Neural network
  • Fuzzy
  • Genetic algorithm
  • Intelligent computer

  • Intelligent agent
  • Intelligent DB search
  • Intelligent robot
  • Intelligent form processing
  • Intelligent character recognition