Boosting Kernel Combination for multi-class image categorization

Alexis Lechervy (1) , Philippe-Henri Gosselin (2) , Frédéric Precioso (3)

1 ETIS - UMR 8051 - Equipes Traitement de l'Information et Systèmes
2 MIDI - Multimedia Indexation and Data Integration
3 Laboratoire d'Informatique, Signaux, et Systèmes de Sophia-Antipolis (I3S) / Equipe KEIA

In this paper, we propose a novel algorithm to design multi- class kernel functions based on an iterative combination of weak kernels in a scheme inspired from boosting framework. The method proposed in this article aims at building a new feature where the centroid for each class are optimally lo- cated. We evaluate our method for image categorization by considering a state-of-the-art image database and by compar- ing our results with reference methods. We show that on the Oxford Flower databases our approach achieves better results than previous state-of-the-art methods.

Keywords: Image databases; Machine learning algorithms; Boosting

Domains: Machine Learning [stat.ML]; Multimedia [cs.MM]

Alexis Lechervy, Philippe-Henri Gosselin, Frédéric Precioso. Boosting Kernel Combination for multi-class image categorization. 2012 IEEE International Conference on Image Processing (ICIP), Sep 2012, Orlando, United States. pp.4. ⟨hal-00753156⟩