Pulsed Neural Networks for Image Processing
Abstract
Еhe information streams model of the artificial pulsed neural network to reserch the performance peculiarities in the up-coming physical systems in the vision applications шы зкуыутеув. The information is transmitted between layers of neurons, performing the convolution, by sequences consisting of pulses of the identical amplitude and width. The simplified theoretical model is tested by the system capable to change the inner and outer performance parameters of the pulsed neuron. The presented model of pulsed network tends to output optimal count of pulses due to negative feedback, which introduces the swinging character of the error function. The paper also presents the quantitative evaluation obtained near minimal error when the pulsed neuron has the response linear and exponential characteristics. The work of both neurons are compared by the extracted factors of the polynomial sum. Therefore, the images were convolved with artificial neural network consisting of the optimized pulse neurons. Ill. 8, bibl. 4 (in English; summaries in English, Russian and Lithuanian).
Downloads
Published
How to Cite
Issue
Section
License
Authors retain copyright and grant the journal the right of the first publication with the paper simultaneously licensed under the Creative Commons Attribution 4.0 (CC BY 4.0) licence.
Authors are allowed to enter into separate, additional contractual arrangements for the non-exclusive distribution of the paper published in the journal with an acknowledgement of the initial publication in the journal.
Copyright terms are indicated in the Republic of Lithuania Law on Copyright and Related Rights, Articles 4-37.