Rejection strategy for convolutional neural network by
adaptive topology applied to handwritten digits
recognition
H. Cecotti, and A. Belaïd. Document Analysis and Recognition, 2005. Proceedings.
Eighth International Conference on, 2, page 765--769. (August 2005)
DOI: 10.1109/ICDAR.2005.200
Abstract
In this paper, we propose a rejection strategy for
convolutional neural network models. The purpose of
this work is to adapt the network's topology injunction
of the geometrical error. A self-organizing map is used
to change the links between the layers leading to a
geometric image transformation occurring directly
inside the network. Instead of learning all the
possible deformation of a pattern, ambiguous patterns
are rejected and the network's topology is modified in
function of their geometric errors thanks to a
specialized self-organizing map. Our objective is to
show how an adaptive topology, without a new learning,
can improve the recognition of rejected patterns in the
case of handwritten digits.
%0 Conference Paper
%1 cecotti-rejection-convolutional-neural-2005
%A Cecotti, Hubert
%A Belaïd, A.bdel
%B Document Analysis and Recognition, 2005. Proceedings.
Eighth International Conference on
%D 2005
%K mnist som
%P 765--769
%R 10.1109/ICDAR.2005.200
%T Rejection strategy for convolutional neural network by
adaptive topology applied to handwritten digits
recognition
%U http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=1575648
%V 2
%X In this paper, we propose a rejection strategy for
convolutional neural network models. The purpose of
this work is to adapt the network's topology injunction
of the geometrical error. A self-organizing map is used
to change the links between the layers leading to a
geometric image transformation occurring directly
inside the network. Instead of learning all the
possible deformation of a pattern, ambiguous patterns
are rejected and the network's topology is modified in
function of their geometric errors thanks to a
specialized self-organizing map. Our objective is to
show how an adaptive topology, without a new learning,
can improve the recognition of rejected patterns in the
case of handwritten digits.
@inproceedings{cecotti-rejection-convolutional-neural-2005,
abstract = {In this paper, we propose a rejection strategy for
convolutional neural network models. The purpose of
this work is to adapt the network's topology injunction
of the geometrical error. A self-organizing map is used
to change the links between the layers leading to a
geometric image transformation occurring directly
inside the network. Instead of learning all the
possible deformation of a pattern, ambiguous patterns
are rejected and the network's topology is modified in
function of their geometric errors thanks to a
specialized self-organizing map. Our objective is to
show how an adaptive topology, without a new learning,
can improve the recognition of rejected patterns in the
case of handwritten digits.},
added-at = {2015-10-27T17:11:36.000+0100},
author = {Cecotti, Hubert and Bela{\"i}d, A.bdel},
biburl = {https://www.bibsonomy.org/bibtex/2b320b2ba7b17a945c52ab02b38240bf4/mhwombat},
booktitle = {Document Analysis and Recognition, 2005. Proceedings.
Eighth International Conference on},
doi = {10.1109/ICDAR.2005.200},
interhash = {0a15c133e289c4ba9cf8a2513548ef20},
intrahash = {b320b2ba7b17a945c52ab02b38240bf4},
issn = {1520-5263},
keywords = {mnist som},
month = aug,
pages = {765--769},
timestamp = {2016-07-12T19:25:30.000+0200},
title = {Rejection strategy for convolutional neural network by
adaptive topology applied to handwritten digits
recognition},
url = {http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=1575648},
volume = 2,
year = 2005
}