2017 Fiscal Year Final Research Report
Face-to-face discussion support system using posture as social signal
Project/Area Number |
15K00275
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Multi-year Fund |
Section | 一般 |
Research Field |
Human interface and interaction
|
Research Institution | Tokyo City University (2017) Kagawa University (2015-2016) |
Principal Investigator |
ICHINO JUNKO 東京都市大学, メディア情報学部, 教授 (50452040)
|
Project Period (FY) |
2015-04-01 – 2018-03-31
|
Keywords | 対面議論支援 / 非言語行動 / 姿勢 / 社会的シグナル |
Outline of Final Research Achievements |
Our goal is to estimate social signals such as degree of interest and degree of consent from users' posture for the purpose of supporting face-to- face discussion. We used a pressure sensor to discriminate the posture of the user. Main posture of the people under discussion is a total of 12 postures including four body parts - head (three postures), torso part (three postures), arm part (three postures), leg part (three postures). First, from the pressure distribution information of the time series, we examine the feature quantity effective for posture discrimination and the extraction method of the feature quantity. We next conducted discrimination using SVM using extracted feature quantities. As a result, the discrimination accuracy of the four body parts was 56.6%, 56.5%, 79.0%, and 82.0%, respectively.
|
Free Research Field |
ヒューマンコンピュータインタラクション
|