Much attention has recently been paid on the recognition of graphical objects, like company logos and trademarks. Recognizing these objects facilitates the recognition of document class. Some promising results have been achieved by using Autoassociator-based Artificial Neural Networks (AANN) in the presence of homogeneously distributed noise. However, the performance drops significantly when dealing with spot-noisy logos, where strips or blobs produce a partial obstruction of the pictures. In this paper, we propose a new approach for training AANNs especially conceived for dealing with spot noise. The basic idea is that of introducing a new metrics for assessing the reproduction error in AANNs. The proposed algorithm, which is referred to as Spot-Backpropagation (S-BP), is significantly more robust with respect to spot-noise than classical Euclidean norm-based Backpropagation (BP). Our experimental results are based on a database of 88 real logos that are artificially corrupted by spot-noise.

Cesarini, F., Francesconi, E., Gori, M., Marinai, S., Sheng, J.Q., Soda, G. (1997). A neural-based architecture for spot-noisy logo recognitionProceedings of the Fourth International Conference on Document Analysis and Recognition. In Proceedings of the Fourth International Conference on Document Analysis and Recognition (pp.175-179). Los Alamitos : IEEE [10.1109/ICDAR.1997.619836].

A neural-based architecture for spot-noisy logo recognitionProceedings of the Fourth International Conference on Document Analysis and Recognition

GORI, MARCO;
1997-01-01

Abstract

Much attention has recently been paid on the recognition of graphical objects, like company logos and trademarks. Recognizing these objects facilitates the recognition of document class. Some promising results have been achieved by using Autoassociator-based Artificial Neural Networks (AANN) in the presence of homogeneously distributed noise. However, the performance drops significantly when dealing with spot-noisy logos, where strips or blobs produce a partial obstruction of the pictures. In this paper, we propose a new approach for training AANNs especially conceived for dealing with spot noise. The basic idea is that of introducing a new metrics for assessing the reproduction error in AANNs. The proposed algorithm, which is referred to as Spot-Backpropagation (S-BP), is significantly more robust with respect to spot-noise than classical Euclidean norm-based Backpropagation (BP). Our experimental results are based on a database of 88 real logos that are artificially corrupted by spot-noise.
1997
0818678984
Cesarini, F., Francesconi, E., Gori, M., Marinai, S., Sheng, J.Q., Soda, G. (1997). A neural-based architecture for spot-noisy logo recognitionProceedings of the Fourth International Conference on Document Analysis and Recognition. In Proceedings of the Fourth International Conference on Document Analysis and Recognition (pp.175-179). Los Alamitos : IEEE [10.1109/ICDAR.1997.619836].
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11365/37854
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo