Abstract
Binary weights are favored in electronic and optical hardware implementations of neural networks as they lead to improved system speeds. Optical neural networks based on fast ferroelectric liquid crystal binary level devices can benefit from the many orders of magnitudes improved liquid crystal response times. An optimized learning algorithm for all-positive perceptrons is simulated on a limited data set of handwritten digits and the resultant network implemented optically. First, gray-scale and then binary inputs and weights are used in recall mode. On comparing the results for the example data set, the binarized inputs and weights network shows almost no loss in performance.
Preview
Unable to display preview. Download preview PDF.
References
S. Abramson, D. Saad, and E. Marom, “Training a Neural Network with Ternary Weights Using the CHIR Algorithm,” IEEE Transactions on Neural Networks, 4(6), 997–1000, (1993).
S. E. Broomfield, M. A. A. Neil, and E. G. S. Paige, “Programmable multiple-level phase modulation that uses ferroelectric liquid crystal spatial light modulation,” Applied Optics, 34(29), 6652–6665 (1995).
N. Collings, A. R. Pourzand, and R. Volkel, “Construction of a programmable multilayer analogue neural network using space invariant interconnects,” in SPIE Vol. 2565, pp. 40–47 (1995).
E. Fiesler, A. Choudry, and H. J. Caulfield, “A Weight Discretization Paradigm for Optical Neural Networks,” in Proceedings of the International Congress on Optical Science and Engineering, vol. SPIE 1281, pp. 164–173, SPIE, Bellingham, Washington (1990).
M. D. Garris and R. A. Wilkinson, data from NIST Special Database 3, National Institute of Standards and Technology (1992).
P. Moerland, E. Fiesler, and I. Saxena, “Incorporation of Liquid-Crystal Light Valve Non-Linearities in Optical Multilayer Neural Networks,” Applied Optics, 35(26), 5301–5307 (1996). P. Moerland, E. Fiesler, and I. Saxena.
P. Moerland, E. Fiesler, and I. Saxena, “Discrete All-Positive Multilayer Perceptrons for Optical Implementations,” IDIAP-RR 97-02, IDIAP, Martigny, Switzerland, February 1997 (submitted to Optical Engineering).
M. G. Robinson and K. M. Johnson, “Noise Analysis of Polarization-Based Optoelectronic Connectionist Machines,” Applied Optics, 31(2), 263–272 (1992).
I. Saxena, P. Moerland, E. Fiesler, A. R. Pourzand, N. Collings, “An Optical Thresholding Perceptron,” in Proceedings of the Workshop on Optics and Computer Science, Geneva, Switzerland (1997).
G. Thimm, P. Moerland, and E. Fiesler, “The Interchangeability of Learning Rate and Gain in Backpropagation Neural Networks,” Neural Computation, 8(2), 451–460 (1996).
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1997 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Saxena, I., Moerland, P., Fiesler, E., Pourzand, A. (1997). Handwritten digit recognition with binary optical perceptron. In: Gerstner, W., Germond, A., Hasler, M., Nicoud, JD. (eds) Artificial Neural Networks — ICANN'97. ICANN 1997. Lecture Notes in Computer Science, vol 1327. Springer, Berlin, Heidelberg. https://doi.org/10.1007/BFb0020323
Download citation
DOI: https://doi.org/10.1007/BFb0020323
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-63631-1
Online ISBN: 978-3-540-69620-9
eBook Packages: Springer Book Archive