2003 Fiscal Year Final Research Report Summary
Design of a face and gesture analysis system that can adapt to changes of user's intention by using anatomical knowledge
| Project/Area Number |
14580417
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | Shizuoka University |
|---|
Principal Investigator |
SAJI Hitoshi Shizuoka University, Faculty of Information, Associate Professor, 情報学部, 助教授 (10283334)
|
|---|
| Project Period (FY) |
2002 – 2003
|
| Keywords | Motion Analysis / 3D Shape Variation / Anatomical Knowledge / Variation of Consciousness / Variation of Facial Expressions |
|---|
| Research Abstract |
In the human facial expressions and gesture, there are intentional and unintentional changes both in space and time. We can realize the human interface that enables face-to-face interaction between a human and a computer by distinguishing the change of human intentions. The purpose of our research is to recognize the human emotions by distinguishing the changes of intention. We studied as follows.
(1)Measurement of facial Shapes To analyze the human facial expressions and gestures that adapt to he change of human, intensions accurately, we have to measure the change of the shape and motion. Therefore, we constructed the system that measures the shape of a human face and the motion of a human gesture. First, we constructed the system that measured the change of the three-dimensional facial shapes.
(2)Tracking of hand motions To analyze the change of the gesture, we constructed the system that tracked the motions of the hand in the three-dimensional space.
(3)Experiments We evaluated our system by experimenting on several subjects, and we considered the difference of the measured values between in the intentional case and the unintentional case.
(4)Design of application system We considered the feature space that can distinguish the human intentions, and we considered the method of analyzing the user's intentions by using anatomical knowledge. Furthermore, we considered the application of our system.
|
