Active learning techniques for user interactive systems: application to image retrieval

Philippe-Henri Gosselin (1) , Matthieu Cord (2)

1 MIDI - Multimedia Indexation and Data Integration
2 ETIS - UMR 8051 - Equipes Traitement de l'Information et Systèmes

Active learning methods have been consid- ered with an increasing interest for user inter- active systems. In this paper, we propose an efficient active learning scheme to deal with this particular context. An active boundary correction is proposed in order to deal with few training data. Experiments are carried out on the COREL photo database.

Domaines: Machine Learning [stat.ML]

Philippe-Henri Gosselin, Matthieu Cord. Active learning techniques for user interactive systems: application to image retrieval. International Workshop on Machine Learning techniques for processing MultiMedia content, Aug 2005, France. pp.1. ⟨hal-00520312⟩