2001.1- "Ubiquitous stereo vision for human sensing"
  To realize an easy and touch-less human interface in a room, we proposed a Ubiquitous Stereo Vision system, which is an extension of the single learning stereo vision system. In this research, our aim was not to reconstruct exact 3D structures or to track objects by active cameras. Instead, by using stereo cameras placed ubiquitously, the system decides on multi-interest regions by using the depth information from each camera, recognizes multi-regions in parallel and integrates all information. Our objective was to create an easy interface and describe all the actions of users. For this purpose, we investigated a method of learning multi-images from multi-cameras in Gigabit Ethernets.

USV

1996.1-2000.12
"Utilization of stereo disparity and optical flow information for human interaction"
  We researched more flexible human interaction by utilizing both stereo disparity and optical flow. To use two data in real time, we utilized multi-resolution images on the Pyramid Vision system.
(main paper)

Demo with a voice (MPEG)!
Speak this side (1.2M) Gestures: User 1 (1.4M)
Gestures: User 2 (1.4M) User Entry/Exit (2.3M)

1992.4-1995.12
"Automatic acquisition of hierarchical mathematical morphology procedures by genetic algorithms"
  We researched a method to obtain an image processing procedure by showing an original image and a target image. We used genetic algorithms (GA) to search procedures. Utilizing a parallel searching function of GA and introducing the concept of era, we constructed hierarchical procedures of mathematical morphology.
(main paper)

Original Image Automatic Acquistion of 
Porcedures by Two Images Goal Image Element of Morphology



2002.11
Doctor Eng. paper: "Real-time recognition system for human-centered scene understanding"
  School of Engineering, University of Tokyo (Paper doctoral degree)
  To attain smooth interaction between a computer and its users, I proposed a system, which obtained depth information and optical flow from stereo gray multi resolution images, and simultaneously used these data to recognize regions nearest to the cameras of the computer. This system identified 1 to 3 nearest regions and recognized objects and their rough movements within the regions. In this method, I utilized higher order local autocorrelation features and linear discriminant analysis to recognize the objects. As the recognition method was a view-based method, the system could handle the face and hand recognitions simultaneously in real-time.

1990.4-1992.3
Master Eng. paper: "Study on visualization of multi-dimensional data by color information"
  Graduate School of Engineering, Tokyo Metropolitan Institute of Technology
  I researched visualization methods for multi-dimensional data by continuously using color information. In the Munsell Color System, colors are arranged in 3D space, consisting of the hue, value and chroma. I attempted to display all the components at the same time by using their properties. This method seemed to improve existing magnetocardiograms and has the advantage of displaying moving images. displaying moving images.
(main paper)

MPEG Demo!
Display of electric charge by Munsell color system (1.1M)

have the five dimensional variables at each sampling point, which are composed of the 2D coordinate and 3D vectors. Therefore it is difficult to display all the components of the vectors simultaneously.
In the method, the Munsell color system is used to represent the vectors.

1989.4-1990.3
Bachelor Eng. paper: "Study on visualization of multi-dimensional data for magnetocardiograms"
  Faculty of Engineering, Tokyo Metropolitan Institute of Technology
  I researched a new visualization method for displaying spatially distributed moving vectors. Some types of scientific data (for example, magnetocardiograms and magnetoencephalograms) have five- dimensional variables at each sampling point, which are composed of the 2D coordinate and 3D vectors, making it difficult to display all the components of the vectors simultaneously. In the method, the Munsell color system is used to represent the vectors.

Magnetocardiogram (926K)
>>> Papers list | Page Top




2005. 4 Assosiate professor, Cooperative Graduate School, Graduate School of Systems & Information Engineering, University of Tsukuba (Additional post)
2005. 4 Group Leader, Ubiquitous Vision Research Group, ITRI, AIST
2004. 7 Senior Research Scientist, Information Technology Research Institute(ITRI), AIST
2002.11 Dr. Eng., University of Tokyo
2001. 4 ETL was transformed into National Institute of Advanced Industrial Science and Technology (AIST)
Senior Research Scientist, Intelligent Systems Institute
2000. 4 Senior Research Scientist, Machine Understanding Div., ETL, MITI
1998. 4 Research Scientist, Adaptive Vision Lab., Machine Understanding Div., ETL
1992. 4 Research Scientist, Image Understanding Sec., Machine Understanding Div.,
Electrotechnical Laboratory (ETL), Ministry of International Trade and Industry (MITI)
1992. 3 M.E., Electronics and Information Systems, Tokyo Metropolitan Institute of Technology
1990. 3 B.E., Information Engineering, Tokyo Metropolitan Institute of Technology
1967 Born in Saku, Nagano, Japan.




2003.7-2006.6 Member of Research Committee: Media Device and Image Information Technology Committee, Institute of Electrical Engineers of Japan (IEEJ)
2002.4-2004.3 Member of Editorial Committee: Journal A, Institute of Electrical Engineers of Japan (IEEJ)
2001.5- Member of Technology Committee: Optical Applications and Vision Committee, Institute of Electrical Engineers of Japan (IEEJ)
2000.7-2003. 6 Member of Research Committee: Media and Image Information Technology Committee, Institute of Electrical Engineers of Japan (IEEJ)
1999.6-2001. 5 Organizer: Audio Visual Information Research Group in Japan (AVIRG)
1999.8-2000. 7 Member of Executive Committee: 6th Symposium on Sensing via Image Information (SSII2000)
1998.8-1999. 7 Member of Executive Committee: 5th symposium on Sensing via Image Information (SII'99)
1997.9-1997.12 Member of Program Committee: Third Symposium on Intelligent Information Media
1997.8-1998. 7 Member of Executive Committee: 4th symposium on Sensing via Image Information (SII'98)
1996.8-1997. 7 Member of Executive Committee: 3rd symposium on Sensing via Image Information (SII'97)
1995.8-1996. 7 Manager of Executive Committee: 2nd symposium on Sensing via Image Information (SII'96)
1995.4-1998. 3 Member of Liaison Committee: CG and CAD Group, Information Processing Society of Japan (IPSJ)



Institute of Electrical and Electronics Engineers (IEEE)
Institute of Electronics, Information and Communication Engineers of Japan (IEICE)
Information Processing Society of Japan (IPSJ)
Institute of Image Information and Television Engineers of Japan (IIITE)
Institute of Electrical Engineers of Japan (IEEJ)
Audio Visual Information Research Group in Japan (AVIRG)


>>> Papers list | Page Top

Copyright© National Institute of Advanced Industrial Science and Technology (AIST). All rights reserved