Authentication keyboard handwriting using a neural network

Main Article Content

Григорий Дмитриевич Асяев
Андрей Николаевич Рагозин

Abstract

In the article, the application of varieties of a neural network for image recognition has been
studied. The main technique of neural network training is revealed. A minimal set of input data
for correct authentication is defined. Experiments were carried out, during which the most
optimal model of a neural network was chosen, which with the least number of errors performed
keyboard-based authentication.

Article Details

Section
Study and design technical means