Neural network architecture for robot hand control

Huan Liu, Thea Iberall, George A. Bekey

Research output: Contribution to journalArticle

24 Citations (Scopus)

Abstract

A robot hand control system called GeSAM, which is under development at the University of Southern California, is described. The goal of the GeSAM architecture is to provide a generic robot hand controller that is based on a model of human prehensile function. It focuses on the relationship between geometric object primitives and the ways a hand can perform prehensile behaviors. It is shown how the relationship between object primitives and a useful set of grasp modes can be learned by an adaptive neural network. By adding training points as necessary, system performance can be improved, avoiding the tedious job of computing every relationship individually.

Original languageEnglish (US)
Pages (from-to)38-43
Number of pages6
JournalIEEE Control Systems Magazine
Volume9
Issue number3
StatePublished - Apr 1989
Externally publishedYes

Fingerprint

End effectors
Network Architecture
Network architecture
Robot
Robots
Neural Networks
Neural networks
Geometric object
Control systems
Controllers
System Performance
Control System
Controller
Necessary
Computing
Relationships
Model

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Neural network architecture for robot hand control. / Liu, Huan; Iberall, Thea; Bekey, George A.

In: IEEE Control Systems Magazine, Vol. 9, No. 3, 04.1989, p. 38-43.

Research output: Contribution to journalArticle

Liu, H, Iberall, T & Bekey, GA 1989, 'Neural network architecture for robot hand control', IEEE Control Systems Magazine, vol. 9, no. 3, pp. 38-43.
Liu, Huan ; Iberall, Thea ; Bekey, George A. / Neural network architecture for robot hand control. In: IEEE Control Systems Magazine. 1989 ; Vol. 9, No. 3. pp. 38-43.
@article{31028e47cf754f1fb64646209e56c8b1,
title = "Neural network architecture for robot hand control",
abstract = "A robot hand control system called GeSAM, which is under development at the University of Southern California, is described. The goal of the GeSAM architecture is to provide a generic robot hand controller that is based on a model of human prehensile function. It focuses on the relationship between geometric object primitives and the ways a hand can perform prehensile behaviors. It is shown how the relationship between object primitives and a useful set of grasp modes can be learned by an adaptive neural network. By adding training points as necessary, system performance can be improved, avoiding the tedious job of computing every relationship individually.",
author = "Huan Liu and Thea Iberall and Bekey, {George A.}",
year = "1989",
month = "4",
language = "English (US)",
volume = "9",
pages = "38--43",
journal = "IEEE Control Systems",
issn = "1066-033X",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "3",

}

TY - JOUR

T1 - Neural network architecture for robot hand control

AU - Liu, Huan

AU - Iberall, Thea

AU - Bekey, George A.

PY - 1989/4

Y1 - 1989/4

N2 - A robot hand control system called GeSAM, which is under development at the University of Southern California, is described. The goal of the GeSAM architecture is to provide a generic robot hand controller that is based on a model of human prehensile function. It focuses on the relationship between geometric object primitives and the ways a hand can perform prehensile behaviors. It is shown how the relationship between object primitives and a useful set of grasp modes can be learned by an adaptive neural network. By adding training points as necessary, system performance can be improved, avoiding the tedious job of computing every relationship individually.

AB - A robot hand control system called GeSAM, which is under development at the University of Southern California, is described. The goal of the GeSAM architecture is to provide a generic robot hand controller that is based on a model of human prehensile function. It focuses on the relationship between geometric object primitives and the ways a hand can perform prehensile behaviors. It is shown how the relationship between object primitives and a useful set of grasp modes can be learned by an adaptive neural network. By adding training points as necessary, system performance can be improved, avoiding the tedious job of computing every relationship individually.

UR - http://www.scopus.com/inward/record.url?scp=0024648022&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0024648022&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0024648022

VL - 9

SP - 38

EP - 43

JO - IEEE Control Systems

JF - IEEE Control Systems

SN - 1066-033X

IS - 3

ER -