Presentation 2012-03-09
Fingerspelling Recognition based on AdaBoost with Hand-shape CG images
Kentaro TAJI, Kohei SETOYAMA, Yasuhiro OHKAWA, Nobuko KATO, Akio OKAZAKI, Kazuhiro FUKUI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Fingerspelling is one of useful communication tools for deaf and hard-of-hearing people. Learning fingerspellings by oneself is often difficult without an experience of using them in real conversation, as learning a foreign language. In the system of fingerspellings, each alphabetical character is expressed with a different hand-shape. Therefore, how to construct a hand-shape recognition method is considered as a central issue to construct an assistant system for learning fingerspellings. The construction of such the recognition system generally requires huge training data set. To avoid this heavy work of data collections, we propose a framework of system learning based on ensemble learning with synthetic images generated by computer graphics techniques. Experimental results show the validity of the proposed framework.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) fingerspelling / hand-shape recognition / pattern recognition / ensemble learning / computer graphics
Paper # WIT2011-71
Date of Issue

Conference Information
Committee WIT
Conference Date 2012/3/2(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Well-being Information Technology(WIT)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Fingerspelling Recognition based on AdaBoost with Hand-shape CG images
Sub Title (in English)
Keyword(1) fingerspelling
Keyword(2) hand-shape recognition
Keyword(3) pattern recognition
Keyword(4) ensemble learning
Keyword(5) computer graphics
1st Author's Name Kentaro TAJI
1st Author's Affiliation University of Tsukuba()
2nd Author's Name Kohei SETOYAMA
2nd Author's Affiliation Tsukuba University of Technology
3rd Author's Name Yasuhiro OHKAWA
3rd Author's Affiliation University of Tsukuba
4th Author's Name Nobuko KATO
4th Author's Affiliation Tsukuba University of Technology
5th Author's Name Akio OKAZAKI
5th Author's Affiliation Tsukuba University of Technology
6th Author's Name Kazuhiro FUKUI
6th Author's Affiliation University of Tsukuba
Date 2012-03-09
Paper # WIT2011-71
Volume (vol) vol.111
Number (no) 472
Page pp.pp.-
#Pages 6
Date of Issue