東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Modeling the mirror: Grasp learning...

Oztop, Erhan.

Linked to FindBook

Google Book

Amazon

博客來

Modeling the mirror: Grasp learning and action recognition.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Modeling the mirror: Grasp learning and action recognition./
Author:	Oztop, Erhan.
Description:	306 p.
Notes:	Source: Dissertation Abstracts International, Volume: 64-06, Section: B, page: 2757.
Contained By:	Dissertation Abstracts International64-06B.
Subject:	Computer Science. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3094363

Modeling the mirror: Grasp learning and action recognition.
Oztop, Erhan.

Modeling the mirror: Grasp learning and action recognition. - 306 p.

Source: Dissertation Abstracts International, Volume: 64-06, Section: B, page: 2757.

Thesis (Ph.D.)--University of Southern California, 2002.

Mirror neurons within a monkey's premotor area F5 fire not only when the monkey performs a certain class of actions but also when the monkey observes another monkey (or the experimenter) perform a similar action. This dissertation presents computational models of the mechanisms involved in learning-to-grasp and action recognition (mirror neurons) in the primate. The hypothesis behind the Mirror Neuron System (MNS) is that it learns to recognize actions of others by self-observation. The Learning to Grasp Model (LGM) complements MNS model by providing the mechanism to learn grasping behavior starting from an elemental set of behaviors.Subjects--Topical Terms:

626642
Computer Science.

Modeling the mirror: Grasp learning and action recognition.
LDR:03119nmm 2200313 4500 001 1865963
005 20041220103641.5
008 130614s2002 eng d
035 $a (UnM)AAI3094363
035 $a AAI3094363
040 $a UnM $c UnM
100 1 $a Oztop, Erhan. $3 1953382
245 1 0 $a Modeling the mirror: Grasp learning and action recognition.
300 $a 306 p.
500 $a Source: Dissertation Abstracts International, Volume: 64-06, Section: B, page: 2757.
500 $a Adviser: Michael A. Arbib.
502 $a Thesis (Ph.D.)--University of Southern California, 2002.
520 $a Mirror neurons within a monkey's premotor area F5 fire not only when the monkey performs a certain class of actions but also when the monkey observes another monkey (or the experimenter) perform a similar action. This dissertation presents computational models of the mechanisms involved in learning-to-grasp and action recognition (mirror neurons) in the primate. The hypothesis behind the Mirror Neuron System (MNS) is that it learns to recognize actions of others by self-observation. The Learning to Grasp Model (LGM) complements MNS model by providing the mechanism to learn grasping behavior starting from an elemental set of behaviors.
520 $a From a computational point of view, LGM is a neural agent that interacts with objects to acquire feasible sets of grasp plans, which are represented as input dependent probability distributions. LGM is adapted using reinforcement learning based on the grasp success. MNS is made up of Core Mirror Circuit and supportive visual and motor schemas. The Visual Analysis schema is implemented as a computer vision system, the Reach and Grasp Simulator schema is implemented as a 3D kinematics arm/hand, and the Core Mirror Circuit schema is implemented as a neural network, which is adapted using supervised learning.
520 $a Both models replicate a large body of existing experimental data and offer explanations to the operational principles underlying their function. Moreover, through simulation experiments, explicit predictions are derived, which can be verified experimentally.
520 $a To be specific, MNS model predicts temporal activity of mirror neurons during the conditions of ambiguous grasp, spatial perturbation and altered kinematics. The analyses further yields insights into the temporal effects of explicit affordance coding, and the compatibility of the hand configuration and movement to object affordance. LGM simulations show that infants can shape their reaching movements into voluntary grasping via goal directed trial and error learning. Furthermore, the model predicts that the task constraints faced during learning are influential in shaping the grasp repertoire of infants and can mediate the learning of precision grasping in early infancy.
590 $a School code: 0208.
650 4 $a Computer Science. $3 626642
650 4 $a Biology, Neuroscience. $3 1017680
650 4 $a Psychology, Developmental. $3 1017557
690 $a 0984
690 $a 0317
690 $a 0620
710 2 0 $a University of Southern California. $3 700129
773 0 $t Dissertation Abstracts International $g 64-06B.
790 1 0 $a Arbib, Michael A., $e advisor
790 $a 0208
791 $a Ph.D.
792 $a 2002
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3094363