Placeholder

Testing artificial neural network for hand gesture recognition


Jòzsef SütÅ‘Ștefan Oniga


Abstract

creative_2013_22_2_223_228_abstract

Neural networks are well applicable for gesture recognition. In this article we want to present the result of an artificial feed forward network for a simplified hand gesture recognition problem. In neural networks, the learning algorithm is very important because the performance of neural network depends on it. One of the most known learning algorithm is the backpropagation. There are some mathematical software which provides acceptable result for a given problem by a backpropagation based network, but in some cases a high-level programming language implemented program can provide better solution. The main topics of the article cover the structure of the test environment, the mathematical background of the implemented methods, some programming remarks and the test results.

Additional Information

Author(s)

Oniga, Ștefan, SütÅ‘, Jòzsef