A dynamically trained neural network for machine vision applications

• Autorzy: Doulamis, N. , Doulamis, A. , Kollias, S.
• Konferencja: Conference on Computer Graphics and Image Processing, GKPO' 98 (5 ; 18-22.05.1998 ; Borki, Poland)
• Języki publikacji: EN

Abstrakty

EN

A dynamically trained neural network is proposed in this paper proper for adapting the network performance to non stationary image or video inputs. The sheme includes, on one hand, a retrieval mechanism which selects the most approprite network from the system memory and, on the other hand, a weight perturbation procedure which adapts the network weights to the current condition. If no suitable network exists in memory, new weights and network structure are crated and them stored for future use. Experimental results are provided indicating the good performance of the proposed system to computer and machine vision applications.

Słowa kluczowe

EN

neural networks; image processing; adaptive training; non stationary inputs

• Czasopismo: Machine Graphics and Vision
• Rocznik: 1998
• Tom: Vol. 7, No. 1/2
• Strony: 113-123
• Opis fizyczny: Bibliogr. 16 poz.

Twórcy

autor

Doulamis, N.

autor

Doulamis, A.

autor

Kollias, S.

• National Technical University of Athens, Electrical and Computer Engineering, Zografou 15773, Athens, Greece,