Advanced

Touch Perception with SOM, Growing Cell Structures and Growing Grids

Johnsson, Magnus LU ; Gil Mendez, David and Balkenius, Christian LU (2008) Towards Autonomous Robotic Systems 2008 p.79-85
Abstract
We have implemented four bio-inspired selforganizing

haptic systems based on proprioception on a 12

d.o.f. anthropomorphic robot hand. The four systems differ in

the kind of self-organizing neural network used for clustering.

For the mapping of the explored objects, one system uses a Self-

Organizing Map (SOM), one uses a Growing Cell Structure

(GCS), one uses a Growing Cell Structure with Deletion of

Neurons (GCS-DN) and one uses a Growing Grid (GG). The

systems were trained and tested with 10 different objects

of different sizes from two different shape categories. The

generalization abilities of the systems were tested with 6 new

objects.... (More)
We have implemented four bio-inspired selforganizing

haptic systems based on proprioception on a 12

d.o.f. anthropomorphic robot hand. The four systems differ in

the kind of self-organizing neural network used for clustering.

For the mapping of the explored objects, one system uses a Self-

Organizing Map (SOM), one uses a Growing Cell Structure

(GCS), one uses a Growing Cell Structure with Deletion of

Neurons (GCS-DN) and one uses a Growing Grid (GG). The

systems were trained and tested with 10 different objects

of different sizes from two different shape categories. The

generalization abilities of the systems were tested with 6 new

objects. The systems showed good performance with the objects

from both the training set as well as in the generalization

experiments, i.e. they mapped the objects according to shape

and size and discriminated individual objects. The GCS-DN

system managed to evolve disconnected networks representing

different clusters in the input space (small cylinders, large

cylinders, small blocks, large blocks), and the generalization

samples activated neurons in a proper subnetwork in all but

one case. (Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Contribution to conference
publication status
published
subject
pages
7 pages
conference name
Towards Autonomous Robotic Systems 2008
language
English
LU publication?
yes
id
b3fced41-2b17-4e23-a44e-78fd3cbd22ef (old id 1288021)
date added to LUP
2009-01-30 09:15:02
date last changed
2016-04-16 11:39:37
@misc{b3fced41-2b17-4e23-a44e-78fd3cbd22ef,
  abstract     = {We have implemented four bio-inspired selforganizing<br/><br>
haptic systems based on proprioception on a 12<br/><br>
d.o.f. anthropomorphic robot hand. The four systems differ in<br/><br>
the kind of self-organizing neural network used for clustering.<br/><br>
For the mapping of the explored objects, one system uses a Self-<br/><br>
Organizing Map (SOM), one uses a Growing Cell Structure<br/><br>
(GCS), one uses a Growing Cell Structure with Deletion of<br/><br>
Neurons (GCS-DN) and one uses a Growing Grid (GG). The<br/><br>
systems were trained and tested with 10 different objects<br/><br>
of different sizes from two different shape categories. The<br/><br>
generalization abilities of the systems were tested with 6 new<br/><br>
objects. The systems showed good performance with the objects<br/><br>
from both the training set as well as in the generalization<br/><br>
experiments, i.e. they mapped the objects according to shape<br/><br>
and size and discriminated individual objects. The GCS-DN<br/><br>
system managed to evolve disconnected networks representing<br/><br>
different clusters in the input space (small cylinders, large<br/><br>
cylinders, small blocks, large blocks), and the generalization<br/><br>
samples activated neurons in a proper subnetwork in all but<br/><br>
one case.},
  author       = {Johnsson, Magnus and Gil Mendez, David and Balkenius, Christian},
  language     = {eng},
  pages        = {79--85},
  title        = {Touch Perception with SOM, Growing Cell Structures and Growing Grids},
  year         = {2008},
}