CONF Saxena-97.3/IDIAP Handwritten Digit Recognition with Binary Optical Perceptron Saxena, Indu Moerland, Perry Fiesler, Emile Pourzand, A. R. Gerstner, W. Ed. Germond, A. Ed. Hasler, M. Ed. Nicoud, J. -D. Ed. EXTERNAL https://publications.idiap.ch/attachments/reports/1997/rr97-15.pdf PUBLIC https://publications.idiap.ch/index.php/publications/showcite/saxena-97.4 Related documents Proceedings of the International Conference on Artificial Neural Networks (ICANN'97) Lecture Notes in Computer Science 1327 1253-1258 1997 Springer-Verlag Berlin IDIAP-RR 97-15 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 hand-written 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. REPORT Saxena-97.4/IDIAP Handwritten Digit Recognition with Binary Optical Perceptron Saxena, Indu Moerland, Perry Fiesler, Emile Pourzand, A. R. EXTERNAL https://publications.idiap.ch/attachments/reports/1997/rr97-15.pdf PUBLIC Idiap-RR-15-1997 1997 IDIAP Published in ``Proceedings of the International Conference on Artificial Neural Networks (ICANN'97)'' 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 hand-written 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.