|Affiliation||Department of Computer Science and Engineering|
|Fields of Research||Motion Media and Graphics / Image Style Informatics / Digital Lighting & Illumination|
|Degree||Doctor of Enginnering (Osaka University)|
|Academic Societies||ACM / IEEE / IEICE / IPSJ|
Please append "tut.jp" to the end of the address above.
|Laboratory website URL||http://www.val.cs.tut.ac.jp|
|Researcher information URL||Researcher information|
Philosophy of our laboratory is a smart visual media for humanity and future environment.
We are developing a new techology, theory, and systems on the basis of computer graphics, images, and motion and optical informatics, which can be a foundation of digital media for next-generations.
Our ongoing research themes are:
* Image processing and analysis for style features for illustratios, which is utilized for retrieval, perceptual metric, and novel functions that can synthesize or convert their styles, etc., and graphical techniques for controlling illuminations.
* Motion data informatics for analysing and synthesizing human movements and behaviors through machine learning technologies.
Theme1：Motion data informatics
State-of-the-art machine learning technologies are introduced for analyzing and synthesizing human motions using motion capture data that can be collected by measuring actual human movements and behaviors. This project develops a novel system for training of dance, sports, skill-training and discriminating subtle human motions, etc..
Selected publications and works
1. "Efficient dynamic skinning with low-rank helper bone controllers", ACM Transactions on Graphics 35(4) 36, SIGGRAPH 2016
2. "Generating Concise Rules for Human Motion Retrieval", Transactions on Information and Systems of The Institute of Electronics, Information and Communication Engineers E93-D(6) 1636-1643 (2010)
3. "Geostatistical Motion Interpolation", ACM Transactions on Graphics 24(3) 1062-1070, SIGGRAPH 2005
4. "Psychological Model for Animating Crowded Pedestrians", Computer Animation & Virtual Worlds, 16, 343-351 (2005)
5. "Motion Map: Image-based Retrieval and Segmentation of Motion Data", ACM SIGGRAPH / Eurographics Symposium on Computer Animation (SCA2004)
Theme2：Image and Graphics for Illustrations and Illuminations
1. Style based illustrative image analysis and conversion
A new horizon of image technology is explored for focusing their 'Style', by introducing state-of-the-art methodology of feature detection, machine learning, and cognitive sciences, while focusing on illustration images. Ongoing project includes style-oriented ranking of clipart images through machine learning, interactive retrieval via style analysis, and auto-synthesis of various fonts using style features, etc.
2. Graphics for Illumination
Smart control technology of LED lightings and illuminations. Ongoing projects include authentic color restoration of lighting scenes, smart image mapping and interactive techniques with a smartphone for illuminations of arrayed LEDs, and robotic controls for optimizing lighting conditions of captured images.
Selected publications and works
1. "An Unsupervised Approach for Comparing Styles of Illustrations", 13th International Workshop on Content-Based Multimedia Indexing (CBMI), 2015 [Best Paper Award]
2. "Distributed LED Lighting System Integrated with Wireless Sensor Network", International Symposium on Technology for Sustainability, 2012
3. "Restoration of color appearance by combining local adaptations for HDR images", AIC2015
Title of class
Programming / Image Processing / Data Mining and Visualization
Others (Awards, Committees, Board members)
Best Paper Award: 13th International Workshop on Content-Based Multimedia Indexing (2015)