Presentation | 2002/3/11 Comparison of Sensory Motion in the Learning of Capturing Task of a Moving Object Shin'ichi MAEHARA, Masanori SUGISAKA, Katsunari SHIBATA, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Acquisition of actions based on prediction, which needs some context information, is important for a robot in dynamic environment. In ease that both target and robot move together, the robot must compensate its motion because the sensory signals are influenced by the robot motion. It is considered that sensory motion is useful to avoid missing a moving object and to know its motion easily. In this paper, a capturing task of a moving object is employed as an environment, in which both robot and target move. Here, the appropriate actions for this task are learned based on the combination of Elman-type recurrent neural network and reinforcement learning. In this paper, the effect of the sensory motions is focused on. Three kinds of sensory motions. 1) looking to a constant direction in absolute coordinates, 2) keeping the object in the center, and 3) fixed on the robot, are employed, and the learning results are compared. Simulation result is shown that the robot obtained the appropriate actions faster in the case of 1) and 2) than 3). |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Prediction / Sensory motion / Forestall action / Reinforcement learning / Recurrent Neural Network |
Paper # | NC2001-153 |
Date of Issue |
Conference Information | |
Committee | NC |
---|---|
Conference Date | 2002/3/11(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | |
Vice Chair | |
Secretary | |
Assistant |
Paper Information | |
Registration To | Neurocomputing (NC) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Comparison of Sensory Motion in the Learning of Capturing Task of a Moving Object |
Sub Title (in English) | |
Keyword(1) | Prediction |
Keyword(2) | Sensory motion |
Keyword(3) | Forestall action |
Keyword(4) | Reinforcement learning |
Keyword(5) | Recurrent Neural Network |
1st Author's Name | Shin'ichi MAEHARA |
1st Author's Affiliation | Department of Electrical and Electronic Engineering. Oita University() |
2nd Author's Name | Masanori SUGISAKA |
2nd Author's Affiliation | Department of Electrical and Electronic Engineering. Oita University |
3rd Author's Name | Katsunari SHIBATA |
3rd Author's Affiliation | Department of Electrical and Electronic Engineering. Oita University |
Date | 2002/3/11 |
Paper # | NC2001-153 |
Volume (vol) | vol.101 |
Number (no) | 735 |
Page | pp.pp.- |
#Pages | 8 |
Date of Issue |