In this paper we promote the idea of using pixel-based models not only for low level vision, but also to extract high level symbolic representations. We use a deep architecture which has the distinctive property of relying on computational units that incorporate classic computer vision invariances and, especially, the scale invariance. The learning algorithm that is proposed, which is based on information theory principles, develops the parameters of the computational units and, at the same time, makes it possible to detect the optimal scale for each pixel. We give experimental evidence of the mechanism of feature extraction at the first level of the hierarchy, which is very much related to SIFT-like features. The comparison shows clearly that, whenever we can rely on the massive availability of training data, the proposed model leads to better performances with respect to SIFT.

Information Theoretic Learning for Pixel-Based Visual Agents / Gori M.; Melacci S.; Lippi M.; Maggini M.. - ELETTRONICO. - 7577:(2012), pp. 864-875. (Intervento presentato al convegno 12th European Conference on Computer Vision tenutosi a Firenze, Italy nel October 7-13, 2012) [10.1007/978-3-642-33783-3_62].

Information Theoretic Learning for Pixel-Based Visual Agents

LIPPI, MARCO;
2012

Abstract

In this paper we promote the idea of using pixel-based models not only for low level vision, but also to extract high level symbolic representations. We use a deep architecture which has the distinctive property of relying on computational units that incorporate classic computer vision invariances and, especially, the scale invariance. The learning algorithm that is proposed, which is based on information theory principles, develops the parameters of the computational units and, at the same time, makes it possible to detect the optimal scale for each pixel. We give experimental evidence of the mechanism of feature extraction at the first level of the hierarchy, which is very much related to SIFT-like features. The comparison shows clearly that, whenever we can rely on the massive availability of training data, the proposed model leads to better performances with respect to SIFT.
2012
Computer Vision -- ECCV 2012
864
875
Information Theoretic Learning for Pixel-Based Visual Agents / Gori M.; Melacci S.; Lippi M.; Maggini M.. - ELETTRONICO. - 7577:(2012), pp. 864-875. (Intervento presentato al convegno 12th European Conference on Computer Vision tenutosi a Firenze, Italy nel October 7-13, 2012) [10.1007/978-3-642-33783-3_62].
Gori M.; Melacci S.; Lippi M.; Maggini M.
File in questo prodotto:
Eventuali allegati, non sono esposti

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/11585/394761
 Attenzione

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

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 8
  • ???jsp.display-item.citation.isi??? 7
social impact