Generation of Intermediate Representation of Shape based on Figure-Ground Segregation
| Project/Area Number |
17H01754
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Cognitive science
|
|---|
| Research Institution | University of Tsukuba |
|---|
Principal Investigator |
Sakai Ko 筑波大学, システム情報系, 教授 (80281666)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
田村 弘 大阪大学, 生命機能研究科, 准教授 (80304038)
山根 ゆか子 大阪大学, 生命機能研究科, 特別研究員(RPD) (70565043)
|
|---|
| Project Period (FY) |
2017-04-01 – 2020-03-31
|
| Project Status |
Completed (Fiscal Year 2019)
|
| Budget Amount *help |
¥17,680,000 (Direct Cost: ¥13,600,000、Indirect Cost: ¥4,080,000)
Fiscal Year 2019: ¥3,510,000 (Direct Cost: ¥2,700,000、Indirect Cost: ¥810,000)
Fiscal Year 2018: ¥6,370,000 (Direct Cost: ¥4,900,000、Indirect Cost: ¥1,470,000)
Fiscal Year 2017: ¥7,800,000 (Direct Cost: ¥6,000,000、Indirect Cost: ¥1,800,000)
|
|---|
| Keywords | 認知科学 / 神経科学 / 実験系心理学 / 情報工学 / 視覚科学 |
|---|
| Outline of Final Research Achievements |
The function of vision is to understand where and what are objects. The images of the outside world are projected onto the retina. The visual cortices need to determine which regions in the retinal image are what objects, which is so called figure-ground (FG) segregation. If the brain failed to correctly segregate figure from ground, it directly led to the incorrect recognition of the object. The present study investigated the cortical mechanisms that perform the construction of object shapes by FG segregation. The study suggested that the perception of shape is constructed through multiple steps, and that the intermediate-level representation of object shape is constructed in the cortical area V4. Moreover, the study suggested that a few tens of neurons represent object shape in a distributed manner.
|
| Academic Significance and Societal Importance of the Research Achievements |
日常的に見る光景は千差万別であり，移動とともに時々刻々と変化する。しかし私達は，初めて見る光景や物体であっても，図地分離に失敗して形状を誤って知覚することは滅多にない。これは，皮質における図地分離の頑健性と秀逸性を示している。視覚皮質のメカニズムは，今日のAIブームの立役者である Deep Learning として応用され，画期的な成果を出している。しかし，脳が実際にどのような計算をしているのかは，まだ殆ど判っておらず，未知の原理・数学体系があるものと考えられている。本研究で明らかにした中間表現の形成・群表現は，脳の計算原理の一端を明らかにしたものである。
|
Report
(4 results)
Research Products
(47 results)
