| Co-Investigator(Kenkyū-buntansha) |
田中 文英 筑波大学, システム情報系, 准教授 (50512787)
尾形 哲也 早稲田大学, 理工学術院, 教授 (00318768)
吉川 雄一郎 大阪大学, 基礎工学研究科, 准教授 (60418530)
西出 俊 徳島大学, ソシオテクノサイエンス研究部, 講師 (30613400)
|
|---|
| Budget Amount *help |
¥249,210,000 (Direct Cost: ¥191,700,000、Indirect Cost: ¥57,510,000)
Fiscal Year 2016: ¥44,200,000 (Direct Cost: ¥34,000,000、Indirect Cost: ¥10,200,000)
Fiscal Year 2015: ¥56,160,000 (Direct Cost: ¥43,200,000、Indirect Cost: ¥12,960,000)
Fiscal Year 2014: ¥46,020,000 (Direct Cost: ¥35,400,000、Indirect Cost: ¥10,620,000)
Fiscal Year 2013: ¥48,880,000 (Direct Cost: ¥37,600,000、Indirect Cost: ¥11,280,000)
Fiscal Year 2012: ¥53,950,000 (Direct Cost: ¥41,500,000、Indirect Cost: ¥12,450,000)
|
|---|
| Outline of Final Research Achievements |
This study aims at understanding the underlying mechanism for social cognitive development and at designing assistant systems for developmental disorders. We suggest that predictive learning of sensorimotor signals plays a key role in cognitive development. First, our computational analysis of infant-caregiver interaction revealed improvement in bodily and social contingency in infants, which is supported by adaptive scaffolding by caregivers. Then, computational modeling of predictive learning investigated to what extent the theory accounts for development. Our robot experiments demonstrated successful development of various abilities as well as its disorders caused by modifications in model parameters. Finally, we designed assistant systems for developmental disorders. Their atypical perception and action were analyzed and reproduced to investigate the influence on social difficulties. There results contribute to better and deeper understanding of the principle of development.
|
