Discrimination of visual pedestrians data by combining projection and prediction learning

Mathieu Lefort 1, 2 Alexander Gepperth 1, 2
1 Flowers - Flowing Epigenetic Robots and Systems
Inria Bordeaux - Sud-Ouest, U2IS - Unité d'Informatique et d'Ingénierie des Systèmes
Abstract : PROPRE is a generic and semi-supervised neural learning paradigm that extracts meaningful concepts of multimodal data flows based on predictability across modalities. It consists on the combination of two computational paradigms. First, a topological projection of each data flow on a self-organizing map (SOM) to reduce input dimension. Second, each SOM activity is used to predict activities in all other SOMs. Predictability measure, that compares predicted and real activities, is used to modulate the SOM learning to favor mutually predictable stimuli. In this article, we study PROPRE applied to a classical visual pedestrian data classification task. The SOM learning modulation introduced in PROPRE improves significantly classification performance.
Type de document :
Communication dans un congrès
ESANN - European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, Apr 2014, Bruges, Belgium. 2014
Liste complète des métadonnées

Littérature citée [17 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-01061654
Contributeur : Mathieu Lefort <>
Soumis le : lundi 8 septembre 2014 - 10:45:31
Dernière modification le : vendredi 8 décembre 2017 - 14:42:15
Document(s) archivé(s) le : mardi 9 décembre 2014 - 11:45:41

Fichier

article.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01061654, version 1

Collections

Citation

Mathieu Lefort, Alexander Gepperth. Discrimination of visual pedestrians data by combining projection and prediction learning. ESANN - European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, Apr 2014, Bruges, Belgium. 2014. 〈hal-01061654〉

Partager

Métriques

Consultations de la notice

200

Téléchargements de fichiers

167