Presentation | 2017-12-07 A study on an indexing method of continuous JSL fingerspelling with moving action alphabets using motion features masaya kato, Masashi morimoto, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | The promotion of sign language is important for the communication with hearing-impaired people.Fingerspelling is one of sign language and easy to learn, and we have proposed an application with the motion sensor for effective learning of fingerspelling.In order to improve the recognition ability for continuous fingerspelling in this application, it is necessary to estimate the moving action of hand and fingers as finger alphabets or transition of alphabets (non alphabet). This paper proposes an indexing method of finger alphabet and non alphabet for sequential motion data obtained from motion sensor. Two motion features are extracted from the motion data, and two-class classification is performed by machine learning.Section estimation of alphabets and non alphabets (indexing) is performed after the classification by scores generated from frame length. The performance experiment of the classification and section estimation for finger alphabet groups show the effectiveness and reveal challenges of the proposed method. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | sign language / fingerspelling / motion features / machine learning / indexing |
Paper # | WIT2017-60 |
Date of Issue | 2017-11-29 (WIT) |
Conference Information | |
Committee | WIT / HI-SIGACI |
---|---|
Conference Date | 2017/12/6(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | AIST Tokyo Waterfront |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | Well-being Information Technology |
Chair | Chikamune Wada(Kyushu Inst. of Tech.) / 小林 真(筑波技大) |
Vice Chair | Daisuke Wakatsuki(Tsukuba Univ. of Tech.) / 河野 純大(筑波技大) |
Secretary | Daisuke Wakatsuki(Nagoya Inst. of Tech.) / 河野 純大(AIST) |
Assistant | Takeaki Shionome(*) / Manabi Miyagi(Tsukuba Univ. of Tech.) / Takashi Handa(Saitama Industrial Technology Center) |
Paper Information | |
Registration To | Technical Committee on Well-being Information Technology / * |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | A study on an indexing method of continuous JSL fingerspelling with moving action alphabets using motion features |
Sub Title (in English) | |
Keyword(1) | sign language |
Keyword(2) | fingerspelling |
Keyword(3) | motion features |
Keyword(4) | machine learning |
Keyword(5) | indexing |
Keyword(6) | |
1st Author's Name | masaya kato |
1st Author's Affiliation | Aichi Institute of Technology(AIT) |
2nd Author's Name | Masashi morimoto |
2nd Author's Affiliation | Aichi Institute of Technology(AIT) |
Date | 2017-12-07 |
Paper # | WIT2017-60 |
Volume (vol) | vol.117 |
Number (no) | WIT-337 |
Page | pp.pp.115-120(WIT), |
#Pages | 6 |
Date of Issue | 2017-11-29 (WIT) |