State representation with perceptual constancy based on active motion

Authors:
Manabu Gouko;Yuichi Kobayashi
Affiliations:
Department of Electrical and Electronic Engineering, College of Engineering, Nihon University, Koriyama, Fukushima, Japan;Department of Electrical and Electronic Engineering, Tokyo University of Agriculture and Technology, Koganei, Tokyo, Japan
Venue:
ICSR'10 Proceedings of the Second international conference on Social robotics
Year:
2010

Citing 6
Cited 0

Technical Note: \cal Q-Learning

Machine Learning
Purposive behavior acquisition for a real robot by vision-based reinforcement learning

Machine Learning - Special issue on robot learning
Ecological robotics

Adaptive Behavior - Special issue on biologically inspired models of navigation
Self-Organizing Maps

Self-Organizing Maps
Introduction to Reinforcement Learning

Introduction to Reinforcement Learning
Motion sketch: acquisition of visual motion guided behaviors

IJCAI'95 Proceedings of the 14th international joint conference on Artificial intelligence - Volume 1

Quantified Score

Hi-index	0.00

Visualization

Abstract

In a robot system, it is important to consider how the outside environment is expressed as a state using sensor information. In this study, we provide a state representation that can express the sensor output changed by environmental change as the same state. It assumes that sensor outputs are probability distributions, and the distances between the distributions of each sensor's output are used to express a state. To confirm the effectiveness of the proposed state representation, we conducted experiments using a mobile robot. The results confirmed that the proposed representation could recognize similar states using a converted sensor output.