A Self Supervised Bi-directional Neural Network (BDSONN) Architecture for Object Extraction Guided by Beta Activation Function and Adaptive Fuzzy Context Sensitive Thresholding
A multilayer self organizing neural neural network
(MLSONN) architecture for binary object extraction, guided by a beta
activation function and characterized by backpropagation of errors
estimated from the linear indices of fuzziness of the network output
states, is discussed. Since the MLSONN architecture is designed to
operate in a single point fixed/uniform thresholding scenario, it does
not take into cognizance the heterogeneity of image information in
the extraction process. The performance of the MLSONN architecture
with representative values of the threshold parameters of the beta
activation function employed is also studied. A three layer bidirectional
self organizing neural network (BDSONN) architecture
comprising fully connected neurons, for the extraction of objects from
a noisy background and capable of incorporating the underlying image
context heterogeneity through variable and adaptive thresholding,
is proposed in this article. The input layer of the network architecture
represents the fuzzy membership information of the image scene to
be extracted. The second layer (the intermediate layer) and the final
layer (the output layer) of the network architecture deal with the self
supervised object extraction task by bi-directional propagation of the
network states. Each layer except the output layer is connected to the
next layer following a neighborhood based topology. The output layer
neurons are in turn, connected to the intermediate layer following
similar topology, thus forming a counter-propagating architecture
with the intermediate layer. The novelty of the proposed architecture
is that the assignment/updating of the inter-layer connection weights
are done using the relative fuzzy membership values at the constituent
neurons in the different network layers. Another interesting feature
of the network lies in the fact that the processing capabilities of
the intermediate and the output layer neurons are guided by a beta
activation function, which uses image context sensitive adaptive
thresholding arising out of the fuzzy cardinality estimates of the
different network neighborhood fuzzy subsets, rather than resorting to
fixed and single point thresholding. An application of the proposed
architecture for object extraction is demonstrated using a synthetic
and a real life image. The extraction efficiency of the proposed
network architecture is evaluated by a proposed system transfer index
characteristic of the network.
[1] R. C. Gonzalez and P. Wintz, Digital Image Processing, MA: Addison-
Wesley, 1977.
[2] A. Rosenfeld and A. C. Kak, Digital Picture Processing, vol. 1, 2nd eds.,
New York: Academic Press, 1982.
[3] M. P. Ekstrom, ed. Digital Image Processing Techniques, New York:
Academic Press, 1984.
[4] M. R. Banham and A. K. Katsaggelos, Digital image restoration, IEEE
Signal Processing Magazine, vol. 14, no. 2, pp. 24-41, 1997.
[5] M. Egmont-Petersen and T. Arts, "Recognition of radiopaque markers
in X-ray images using a neural network as nonlinear fi lter," Pattern
Recognition Letters, vol. 20, no. 5, pp. 521-533, 1999.
[6] S. Haykin, Neural networks: a comprehensive foundation, Macmillan
College Publishing Co., New York, 1994.
[7] J. Hertz, A. Krogh, R. G. Palmer, Introduction to the theory of neural
computation, Addison-Wesley, 1991.
[8] Y. H. Pao, Adaptive Pattern Recognition and Neural Networks, Addison-
Wesley New York, 1989.
[9] G. L. Bilbro, M. White and W. Synder, "Image segmentation with
neurocomputers," in Neural Computers eds. R. Eckmiller and C. V. D.
Malsburg, Springer-Verlag New York, 1988.
[10] R. P. Lippmann, "An introduction to computing with neural nets," IEEE
ASSP Magazine, pp. 3-22, 1987.
[11] T. Kim, V. Devarajan and M. Manry, "Road extraction from aerial images
using neural networks," Proceedings of ASPRS Annual Convention,
vol. 3, pp. 146-154.
[12] M. Antonucci, B. Tirozzi, N. D. Yarunin et al., "Numerical simulation
of neural networks with translation and rotation invariant pattern recognition,"
International Journal of Modern Physics B, vol. 8, no. 11-12, pp.
1529-1541, 1994.
[13] G. A. Carpenter, and W. D. Ross, "ART-EMAP: A neural network
architecture for object recognition by evidence accumulation," IEEE
Transactions on Neural Networks, vol. 6, no. 4, pp. 805-818, 1995.
[14] T. Kohonen, "Self-organized formation of topologically correct feature
maps," Biological Cybernetics, vol. 43, pp. 59-69, 1982.
15] T. Kohonen, "Self-organization and associative memory," Springer-
Verlag, London, 1984.
[16] J. J. Hopfi eld, "Neurons with graded response have collective computational
properties like those of two state neurons," Proceedings of Nat.
Acad. Sci. U. S. pp. 3088-3092, 1984.
[17] [6] B. Kosko, "Bidirectional associative memories," IEEE Transactions
on Systems, Man and Cybernetics, vol. 18, no. 1, pp. 49-60, 1988.
[18] B. Kosko, Neural Networks and Fuzzy Systems: A Dynamical Systems
Approach to Machine Intelligence, Prentice-Hall Englewood Cliffs NJ,
1992.
[19] W. Chua and l. Yang, "Cellular network: theory," IEEE Transactions on
Circuits and Systems, vol. 35, no. 10, pp. 1257-1282, 1988.
[20] W. Chua and L. Yang, "Cellular network: applications," IEEE Transactions
on Circuits and Systems, vol. 35, no. 10, pp. 1273-1290, 1988.
[21] A. K. Datta, S. Munshi and S. Bhattacharyya, "Object Extraction In
Artifi cial Retina Using Cellular Neural Network Optimized By Genetic
Algorithm With Fuzziness Measure," Proceedings of International Conference
on Fiber Optics and Photonics, vol. 2, pp. 723-725, 2000.
[22] G. A. Carpenter and S. Grossberg, Pattern Recognition by Self-
Organizing Neural Networks, Cambridge, MA: MIT Press, 1991.
[23] A. Ghosh, N. R. Pal, and S. K. Pal, "Self-organization for object
extraction using multilayer neural network and fuzziness measures," IEEE
Transactions on Fuzzy Systems, vol. 1, no. 1, pp. 54-68, 1993.
[24] R. o. Duda and P. E. Hart, Pattern Classifi cation and Scene Analysis,
New York: Wiley, 1973.
[25] J. T. Tou and R. C. Gonzalez, Pattern Recognition Principles, Reading,
MA: Addison Wesley, 1974.
[26] S. Bhattacharyya, P. Dutta and U. Maulik, "Multiple Target Tracking
Using Self-Organizing Neural Network Based Segmentation of Optical
Flow Field," Proceedings of Conference of Recent Trends In Manufacturing,
pp. 369-376, 2003.
[27] A. Ghosh and A. Sen, "Self organizing neural network for multi-level
image segmentation," Soft Computing Applications to Pattern Recognition
and Image Processing, eds. A. Ghosh and S. K. Pal, pp. 129-144, World
Scientifi c, 2002.
[28] S. Bhattacharyya, U. Maulik and S. Bandyopadhyay, "A Fuzzy Cardinality
Based Approximation for Extracting Multi-Scale Objects From
Noisy Background Using Self Organizing Neural Network," Proceedings
of International Conference on Communications, Devices and Intelligent
Systems, pp. 461-464, 2004.
[29] S. Bhattacharyya, P. Dutta and U. Maulik, "Multi-Scale Object Extraction
Using Self Organizing Neural Network With A Multi-Level Sigmoidal
Activation Function," Proceedings of 5th International Conference
on Advances in Pattern Recognition, pp. 435-438, 2003.
[30] P. Dutta, S. Bhattacharyya and K. Dasgupta, "Multi-Scale Object Extraction
Using A Self Organizing Neural Network With A Multi-Level
Beta Activation Function," Proceedings of International Conference on
Intelligent Sensing and Information Processing, pp. 139-142, 2004.
[31] S. Bhattacharyya, P. Dutta and D. DuttaMajumder, "Multiscale Object
Extraction Using A Self-Organizing Neural Network With Multilevel
Beta Activation Function and its Sigmoidal Counterpart: A Comparative
Study," Proceedings of International Conference on Recent Trends and
New Directions of Research in Cybernetics & Systems Theory, 2004.
[32] S. Bhattacharyya and P. Dutta, "Multiscale Object Extraction with
MUSIG and MUBET with CONSENT: A Comparative Study," Proceedings
of KBCS 2004, pp. 100-109, India.
[33] S. Bhattacharyya and P. Dutta, "XMUBET with CONSENT: A Pixel
Hostility Induced Multiscale Object Extractor," Proceedings of IEEE
International Conference on Intelligent Sensors, Sensor Networks and
Information Processing (ISSNIP 04), pp. 277-282, 2004.
[34] S. Bhattacharyya and P. Dutta, "XMUSIG with CONSENT: Pixel Hostility
Induced Multiscale Object Extractor," Proceedings of International
Conference on Intelligent Sensing and Information Processing (ICISIP
2005), 2005.
[35] S. Bhattacharyya and K. Dasgupta, "Color Object Extraction From A
Noisy Background Using Parallel Multi-layer Self-Organizing Neural
Networks," Proceedings of CSI-YITPA(E) 2003, pp. 32-36, 2003.
[36] S. Bhattacharyya and P. Dutta, "A Parallel Self-Organizing Neural Net
work Architecture For Extraction of Graded Color Objects," Proceedings
of International Conference on Recent Trends and New Directions of
Research in Cybernetics & Systems Theory, 2004.
[37] S. Bhattacharyya, P. Dutta and U. Maulik, "Graded color object extraction
by a Parallel Self-Organizing Neural Network (PSONN) architecture
guided by a MUBET activation function," Proceedings of ICIS 2005,
2005.
[38] S. Bhattacharyya, P. Dutta and U. Maulik, "MUSIG and MUBET guided
PSONN based color object extraction: A Comparative Study," Proceedings
of International Conference on Information Technology 2005, 2005.
[39] S. Bhattacharyya, P. Dutta and P. Nandi, "True Color Object Extraction
By A Parallel Self Organizing Neural Network (PSONN) Architecture
Guided By XMUBET With CONSENT," Proceedings of EAIT 2006,
2006.
[40] S. Bhattacharyya and P. Dutta, "Designing pruned multilayer self organizing
neural network (MLSONN) for object extraction from a noisy
background," Book of Abstracts of SOOP 05, pp. 131-132, 2005.
[41] S. Bhattacharyya and P.Dutta, "Designing pruned neighborhood neural
networks for object extraction from noisy background," International
Journal of Foundations of Computing and Decision Sciences, vol. 31,
no. 2, pp. 105-134, 2006.
[42] L. A. Zadeh, "Fuzzy sets," Inform. and Control, vol. 8, no. 1, pp.338-
353, 1965.
[43] T. J. Ross and T. Ross, Fuzzy Logic With Engineering Applications,
McGraw Hill College Div., 1995.
[44] A. Deluca and S. Termini, "A defi nition of non probabilistic entropy
in the setting of fuzzy set theory," Information and Control, vol. 20, pp.
301-312, 1972.
[1] R. C. Gonzalez and P. Wintz, Digital Image Processing, MA: Addison-
Wesley, 1977.
[2] A. Rosenfeld and A. C. Kak, Digital Picture Processing, vol. 1, 2nd eds.,
New York: Academic Press, 1982.
[3] M. P. Ekstrom, ed. Digital Image Processing Techniques, New York:
Academic Press, 1984.
[4] M. R. Banham and A. K. Katsaggelos, Digital image restoration, IEEE
Signal Processing Magazine, vol. 14, no. 2, pp. 24-41, 1997.
[5] M. Egmont-Petersen and T. Arts, "Recognition of radiopaque markers
in X-ray images using a neural network as nonlinear fi lter," Pattern
Recognition Letters, vol. 20, no. 5, pp. 521-533, 1999.
[6] S. Haykin, Neural networks: a comprehensive foundation, Macmillan
College Publishing Co., New York, 1994.
[7] J. Hertz, A. Krogh, R. G. Palmer, Introduction to the theory of neural
computation, Addison-Wesley, 1991.
[8] Y. H. Pao, Adaptive Pattern Recognition and Neural Networks, Addison-
Wesley New York, 1989.
[9] G. L. Bilbro, M. White and W. Synder, "Image segmentation with
neurocomputers," in Neural Computers eds. R. Eckmiller and C. V. D.
Malsburg, Springer-Verlag New York, 1988.
[10] R. P. Lippmann, "An introduction to computing with neural nets," IEEE
ASSP Magazine, pp. 3-22, 1987.
[11] T. Kim, V. Devarajan and M. Manry, "Road extraction from aerial images
using neural networks," Proceedings of ASPRS Annual Convention,
vol. 3, pp. 146-154.
[12] M. Antonucci, B. Tirozzi, N. D. Yarunin et al., "Numerical simulation
of neural networks with translation and rotation invariant pattern recognition,"
International Journal of Modern Physics B, vol. 8, no. 11-12, pp.
1529-1541, 1994.
[13] G. A. Carpenter, and W. D. Ross, "ART-EMAP: A neural network
architecture for object recognition by evidence accumulation," IEEE
Transactions on Neural Networks, vol. 6, no. 4, pp. 805-818, 1995.
[14] T. Kohonen, "Self-organized formation of topologically correct feature
maps," Biological Cybernetics, vol. 43, pp. 59-69, 1982.
15] T. Kohonen, "Self-organization and associative memory," Springer-
Verlag, London, 1984.
[16] J. J. Hopfi eld, "Neurons with graded response have collective computational
properties like those of two state neurons," Proceedings of Nat.
Acad. Sci. U. S. pp. 3088-3092, 1984.
[17] [6] B. Kosko, "Bidirectional associative memories," IEEE Transactions
on Systems, Man and Cybernetics, vol. 18, no. 1, pp. 49-60, 1988.
[18] B. Kosko, Neural Networks and Fuzzy Systems: A Dynamical Systems
Approach to Machine Intelligence, Prentice-Hall Englewood Cliffs NJ,
1992.
[19] W. Chua and l. Yang, "Cellular network: theory," IEEE Transactions on
Circuits and Systems, vol. 35, no. 10, pp. 1257-1282, 1988.
[20] W. Chua and L. Yang, "Cellular network: applications," IEEE Transactions
on Circuits and Systems, vol. 35, no. 10, pp. 1273-1290, 1988.
[21] A. K. Datta, S. Munshi and S. Bhattacharyya, "Object Extraction In
Artifi cial Retina Using Cellular Neural Network Optimized By Genetic
Algorithm With Fuzziness Measure," Proceedings of International Conference
on Fiber Optics and Photonics, vol. 2, pp. 723-725, 2000.
[22] G. A. Carpenter and S. Grossberg, Pattern Recognition by Self-
Organizing Neural Networks, Cambridge, MA: MIT Press, 1991.
[23] A. Ghosh, N. R. Pal, and S. K. Pal, "Self-organization for object
extraction using multilayer neural network and fuzziness measures," IEEE
Transactions on Fuzzy Systems, vol. 1, no. 1, pp. 54-68, 1993.
[24] R. o. Duda and P. E. Hart, Pattern Classifi cation and Scene Analysis,
New York: Wiley, 1973.
[25] J. T. Tou and R. C. Gonzalez, Pattern Recognition Principles, Reading,
MA: Addison Wesley, 1974.
[26] S. Bhattacharyya, P. Dutta and U. Maulik, "Multiple Target Tracking
Using Self-Organizing Neural Network Based Segmentation of Optical
Flow Field," Proceedings of Conference of Recent Trends In Manufacturing,
pp. 369-376, 2003.
[27] A. Ghosh and A. Sen, "Self organizing neural network for multi-level
image segmentation," Soft Computing Applications to Pattern Recognition
and Image Processing, eds. A. Ghosh and S. K. Pal, pp. 129-144, World
Scientifi c, 2002.
[28] S. Bhattacharyya, U. Maulik and S. Bandyopadhyay, "A Fuzzy Cardinality
Based Approximation for Extracting Multi-Scale Objects From
Noisy Background Using Self Organizing Neural Network," Proceedings
of International Conference on Communications, Devices and Intelligent
Systems, pp. 461-464, 2004.
[29] S. Bhattacharyya, P. Dutta and U. Maulik, "Multi-Scale Object Extraction
Using Self Organizing Neural Network With A Multi-Level Sigmoidal
Activation Function," Proceedings of 5th International Conference
on Advances in Pattern Recognition, pp. 435-438, 2003.
[30] P. Dutta, S. Bhattacharyya and K. Dasgupta, "Multi-Scale Object Extraction
Using A Self Organizing Neural Network With A Multi-Level
Beta Activation Function," Proceedings of International Conference on
Intelligent Sensing and Information Processing, pp. 139-142, 2004.
[31] S. Bhattacharyya, P. Dutta and D. DuttaMajumder, "Multiscale Object
Extraction Using A Self-Organizing Neural Network With Multilevel
Beta Activation Function and its Sigmoidal Counterpart: A Comparative
Study," Proceedings of International Conference on Recent Trends and
New Directions of Research in Cybernetics & Systems Theory, 2004.
[32] S. Bhattacharyya and P. Dutta, "Multiscale Object Extraction with
MUSIG and MUBET with CONSENT: A Comparative Study," Proceedings
of KBCS 2004, pp. 100-109, India.
[33] S. Bhattacharyya and P. Dutta, "XMUBET with CONSENT: A Pixel
Hostility Induced Multiscale Object Extractor," Proceedings of IEEE
International Conference on Intelligent Sensors, Sensor Networks and
Information Processing (ISSNIP 04), pp. 277-282, 2004.
[34] S. Bhattacharyya and P. Dutta, "XMUSIG with CONSENT: Pixel Hostility
Induced Multiscale Object Extractor," Proceedings of International
Conference on Intelligent Sensing and Information Processing (ICISIP
2005), 2005.
[35] S. Bhattacharyya and K. Dasgupta, "Color Object Extraction From A
Noisy Background Using Parallel Multi-layer Self-Organizing Neural
Networks," Proceedings of CSI-YITPA(E) 2003, pp. 32-36, 2003.
[36] S. Bhattacharyya and P. Dutta, "A Parallel Self-Organizing Neural Net
work Architecture For Extraction of Graded Color Objects," Proceedings
of International Conference on Recent Trends and New Directions of
Research in Cybernetics & Systems Theory, 2004.
[37] S. Bhattacharyya, P. Dutta and U. Maulik, "Graded color object extraction
by a Parallel Self-Organizing Neural Network (PSONN) architecture
guided by a MUBET activation function," Proceedings of ICIS 2005,
2005.
[38] S. Bhattacharyya, P. Dutta and U. Maulik, "MUSIG and MUBET guided
PSONN based color object extraction: A Comparative Study," Proceedings
of International Conference on Information Technology 2005, 2005.
[39] S. Bhattacharyya, P. Dutta and P. Nandi, "True Color Object Extraction
By A Parallel Self Organizing Neural Network (PSONN) Architecture
Guided By XMUBET With CONSENT," Proceedings of EAIT 2006,
2006.
[40] S. Bhattacharyya and P. Dutta, "Designing pruned multilayer self organizing
neural network (MLSONN) for object extraction from a noisy
background," Book of Abstracts of SOOP 05, pp. 131-132, 2005.
[41] S. Bhattacharyya and P.Dutta, "Designing pruned neighborhood neural
networks for object extraction from noisy background," International
Journal of Foundations of Computing and Decision Sciences, vol. 31,
no. 2, pp. 105-134, 2006.
[42] L. A. Zadeh, "Fuzzy sets," Inform. and Control, vol. 8, no. 1, pp.338-
353, 1965.
[43] T. J. Ross and T. Ross, Fuzzy Logic With Engineering Applications,
McGraw Hill College Div., 1995.
[44] A. Deluca and S. Termini, "A defi nition of non probabilistic entropy
in the setting of fuzzy set theory," Information and Control, vol. 20, pp.
301-312, 1972.
@article{"International Journal of Information, Control and Computer Sciences:62174", author = "Siddhartha Bhattacharyya and Paramartha Dutta and Ujjwal Maulik and Prashanta Kumar Nandi", title = "A Self Supervised Bi-directional Neural Network (BDSONN) Architecture for Object Extraction Guided by Beta Activation Function and Adaptive Fuzzy Context Sensitive Thresholding", abstract = "A multilayer self organizing neural neural network
(MLSONN) architecture for binary object extraction, guided by a beta
activation function and characterized by backpropagation of errors
estimated from the linear indices of fuzziness of the network output
states, is discussed. Since the MLSONN architecture is designed to
operate in a single point fixed/uniform thresholding scenario, it does
not take into cognizance the heterogeneity of image information in
the extraction process. The performance of the MLSONN architecture
with representative values of the threshold parameters of the beta
activation function employed is also studied. A three layer bidirectional
self organizing neural network (BDSONN) architecture
comprising fully connected neurons, for the extraction of objects from
a noisy background and capable of incorporating the underlying image
context heterogeneity through variable and adaptive thresholding,
is proposed in this article. The input layer of the network architecture
represents the fuzzy membership information of the image scene to
be extracted. The second layer (the intermediate layer) and the final
layer (the output layer) of the network architecture deal with the self
supervised object extraction task by bi-directional propagation of the
network states. Each layer except the output layer is connected to the
next layer following a neighborhood based topology. The output layer
neurons are in turn, connected to the intermediate layer following
similar topology, thus forming a counter-propagating architecture
with the intermediate layer. The novelty of the proposed architecture
is that the assignment/updating of the inter-layer connection weights
are done using the relative fuzzy membership values at the constituent
neurons in the different network layers. Another interesting feature
of the network lies in the fact that the processing capabilities of
the intermediate and the output layer neurons are guided by a beta
activation function, which uses image context sensitive adaptive
thresholding arising out of the fuzzy cardinality estimates of the
different network neighborhood fuzzy subsets, rather than resorting to
fixed and single point thresholding. An application of the proposed
architecture for object extraction is demonstrated using a synthetic
and a real life image. The extraction efficiency of the proposed
network architecture is evaluated by a proposed system transfer index
characteristic of the network.", keywords = "Beta activation function, fuzzy cardinality, multilayer self organizing neural network, object extraction,", volume = "2", number = "4", pages = "1267-21", }
