| Project/Area Number |
16300062
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Perception information processing/Intelligent robotics
|
|---|
| Research Institution | National Institute of Advanced Industrial Science and Technology. (AIST) |
|---|
Principal Investigator |
NAKATA Toru National Institute of Advanced Industrial Science and Technology., Digital Human Research Center, Researcher., デジタルヒューマン研究センター, 研究員 (10344152)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
MATUI Toshihiro National Institute of Advanced Industrial Science and Technology., Digital Human Research Center, Managing Researcher., デジタルヒューマン研究センター, 総括研究員 (90358010)
|
|---|
| Project Period (FY) |
2004 – 2005
|
| Project Status |
Completed (Fiscal Year 2005)
|
| Budget Amount *help |
¥12,900,000 (Direct Cost: ¥12,900,000)
Fiscal Year 2005: ¥1,700,000 (Direct Cost: ¥1,700,000)
Fiscal Year 2004: ¥11,200,000 (Direct Cost: ¥11,200,000)
|
|---|
| Keywords | Behavior recognition / Motion recognition / Motion segmentation / Movie understanding / Human body action / モバイル装置 / 動作認識 / ジェスチャ認識 / 身体動作認識 / 認識工学 / 身体運動 / 実世界コンピューティング |
|---|
| Research Abstract |
General human behaviors, i.e. behaviors in daily life, are made as sequese of primitive units of body actions. The primitive actions are called "action alphabet." We can recognize human behavior by comprehending sequence of action alphabets. In order to recognize human behavior by computers, we produced a new method to segment human motion data. The behavior segmentation means chopping motion data into temporal segments of action alphabets. Under our assumption, action alphabets are periods in which condition of body control does not change drastically. We focused on correlations among movements of limbs of a body. The condition of correlations will be maintained during a primitive action segment lasts. In experiment, our system could segment human body motion data into primitive segments. Using these segmentation results, the system succeded to recognize kind of behavior and digest long motion data into small presentation style.
|
