Image Associative Memory by Recurrent Neural Subnetworks

Wfadysfaw SKARBEK; Andrzej CICHOCKI

Image Associative Memory by Recurrent Neural Subnetworks

Wfadysfaw SKARBEK, Andrzej CICHOCKI

Full Text Views

0

Share
Cite this

Summary :

Gray scale images are represented by recurrent neural subnetworks which together with a competition layer create an associative memory. The single recurrent subnetwork N_i implements a stochastic nonlinear fractal operator F_i, constructed for the given image f_i. We show that under realstic assumptions F has a unique attractor which is located in the vicinity of the original image. Therefore one subnetwork represents one original image. The associative recall is implemented in two stages. Firstly, the competition layer finds the most invariant subnetwork for the given input noisy image g. Next, the selected recurrent subnetwork in few (5-10) global iterations produces high quality approximation of the original image. The degree of invariance for the subnetwork N_i on the inprt g is measured by a norm ||g-F_i(g)||. We have experimentally verified that associative recall for images of natural scenes with pixel values in [0, 255] is successful even when Gaussian noise has the standard deviation σ as large as 500. Moreover, the norm, computed only on 10% of pixels chosen randomly from images still successfuly recalls a close approximation of original image. Comparing to Amari-Hopfield associative memory, our solution has no spurious states, is less sensitive to noise, and its network complexity is significantly lower. However, for each new stored image a new subnetwork must be added.

Publication: IEICE TRANSACTIONS on Fundamentals Vol.E79-A No.10 pp.1638-1646

Publication Date: 1996/10/25

Publicized

Online ISSN

DOI

Type of Manuscript: Special Section PAPER (Special Section on Nonlinear Theory and its Applications (NOLTA))

Category: Neural Nets and Human Being

Cite this

Copy

Wfadysfaw SKARBEK, Andrzej CICHOCKI, "Image Associative Memory by Recurrent Neural Subnetworks" in IEICE TRANSACTIONS on Fundamentals, vol. E79-A, no. 10, pp. 1638-1646, October 1996, doi: .
Abstract: Gray scale images are represented by recurrent neural subnetworks which together with a competition layer create an associative memory. The single recurrent subnetwork N_i implements a stochastic nonlinear fractal operator F_i, constructed for the given image f_i. We show that under realstic assumptions F has a unique attractor which is located in the vicinity of the original image. Therefore one subnetwork represents one original image. The associative recall is implemented in two stages. Firstly, the competition layer finds the most invariant subnetwork for the given input noisy image g. Next, the selected recurrent subnetwork in few (5-10) global iterations produces high quality approximation of the original image. The degree of invariance for the subnetwork N_i on the inprt g is measured by a norm ||g-F_i(g)||. We have experimentally verified that associative recall for images of natural scenes with pixel values in [0, 255] is successful even when Gaussian noise has the standard deviation σ as large as 500. Moreover, the norm, computed only on 10% of pixels chosen randomly from images still successfuly recalls a close approximation of original image. Comparing to Amari-Hopfield associative memory, our solution has no spurious states, is less sensitive to noise, and its network complexity is significantly lower. However, for each new stored image a new subnetwork must be added.
URL: https://globals.ieice.org/en_transactions/fundamentals/10.1587/e79-a_10_1638/_p

Copy

@ARTICLE{e79-a_10_1638,
author={Wfadysfaw SKARBEK, Andrzej CICHOCKI, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Image Associative Memory by Recurrent Neural Subnetworks},
year={1996},
volume={E79-A},
number={10},
pages={1638-1646},
abstract={Gray scale images are represented by recurrent neural subnetworks which together with a competition layer create an associative memory. The single recurrent subnetwork N_i implements a stochastic nonlinear fractal operator F_i, constructed for the given image f_i. We show that under realstic assumptions F has a unique attractor which is located in the vicinity of the original image. Therefore one subnetwork represents one original image. The associative recall is implemented in two stages. Firstly, the competition layer finds the most invariant subnetwork for the given input noisy image g. Next, the selected recurrent subnetwork in few (5-10) global iterations produces high quality approximation of the original image. The degree of invariance for the subnetwork N_i on the inprt g is measured by a norm ||g-F_i(g)||. We have experimentally verified that associative recall for images of natural scenes with pixel values in [0, 255] is successful even when Gaussian noise has the standard deviation σ as large as 500. Moreover, the norm, computed only on 10% of pixels chosen randomly from images still successfuly recalls a close approximation of original image. Comparing to Amari-Hopfield associative memory, our solution has no spurious states, is less sensitive to noise, and its network complexity is significantly lower. However, for each new stored image a new subnetwork must be added.},
keywords={},
doi={},
ISSN={},
month={October},}

Copy

TY - JOUR
TI - Image Associative Memory by Recurrent Neural Subnetworks
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 1638
EP - 1646
AU - Wfadysfaw SKARBEK
AU - Andrzej CICHOCKI
PY - 1996
DO -
JO - IEICE TRANSACTIONS on Fundamentals
SN -
VL - E79-A
IS - 10
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - October 1996
AB - Gray scale images are represented by recurrent neural subnetworks which together with a competition layer create an associative memory. The single recurrent subnetwork N_i implements a stochastic nonlinear fractal operator F_i, constructed for the given image f_i. We show that under realstic assumptions F has a unique attractor which is located in the vicinity of the original image. Therefore one subnetwork represents one original image. The associative recall is implemented in two stages. Firstly, the competition layer finds the most invariant subnetwork for the given input noisy image g. Next, the selected recurrent subnetwork in few (5-10) global iterations produces high quality approximation of the original image. The degree of invariance for the subnetwork N_i on the inprt g is measured by a norm ||g-F_i(g)||. We have experimentally verified that associative recall for images of natural scenes with pixel values in [0, 255] is successful even when Gaussian noise has the standard deviation σ as large as 500. Moreover, the norm, computed only on 10% of pixels chosen randomly from images still successfuly recalls a close approximation of original image. Comparing to Amari-Hopfield associative memory, our solution has no spurious states, is less sensitive to noise, and its network complexity is significantly lower. However, for each new stored image a new subnetwork must be added.
ER -