Title :
"Weak rod" by a double layer parallel network
Author :
Matsumoto, T. ; Kondo, K.
Author_Institution :
Dept. of Electr. Eng., Waseda Univ., Tokyo, Japan
Abstract :
Weak rod, a minimisation problem, can be realized by a very natural double layer network with feedback. While the second layer is linear, the first layer is significantly nonlinear. Elements associated with the first layer are called "synaptic fuses". An MRF argument is used to eliminate the line variable. The network makes it transparent what is cut when a crease or a step edge is present.
Keywords :
Markov processes; feedback; minimisation; multilayer perceptrons; MRF argument; Markov random fields; crease; double layer parallel network; feedback; minimisation problem; neural net; step edge; synaptic fuses; weak rod; Algorithm design and analysis; Analog computers; Computer networks; Concurrent computing; Feedback; Fuses; Probability distribution; Random variables; Voltage control; Yagi-Uda antennas;
Conference_Titel :
Neural Networks, 1993. IJCNN '93-Nagoya. Proceedings of 1993 International Joint Conference on
Print_ISBN :
0-7803-1421-2
DOI :
10.1109/IJCNN.1993.714048
