PAC-Bayesian Majority Vote for Late Classifier Fusion

Emilie Morvant (1, 2) , Amaury Habrard (2) , Stéphane Ayache (1)

1 QARMA - éQuipe AppRentissage et MultimediA [Marseille]
2 LHC - Laboratoire Hubert Curien

A lot of attention has been devoted to multimedia indexing over the past few years. In the literature, we often consider two kinds of fusion schemes: The early fusion and the late fusion. In this paper we focus on late classifier fusion, where one combines the scores of each modality at the decision level. To tackle this problem, we investigate a recent and elegant well-founded quadratic program named MinCq coming from the Machine Learning PAC-Bayes theory. MinCq looks for the weighted combination, over a set of real-valued functions seen as voters, leading to the lowest misclassification rate, while making use of the voters' diversity. We provide evidence that this method is naturally adapted to late fusion procedure. We propose an extension of MinCq by adding an order- preserving pairwise loss for ranking, helping to improve Mean Averaged Precision measure. We confirm the good behavior of the MinCq-based fusion approaches with experiments on a real image benchmark.

Mots clés: Machine Learning; Multimedia fusion; Multi-modality search; Ranking and re-ranking

Domaines: Machine Learning [stat.ML]; Multimédia [cs.MM]

Emilie Morvant, Amaury Habrard, Stéphane Ayache. PAC-Bayesian Majority Vote for Late Classifier Fusion. [Research Report] LIF Marseille; LaHC Saint-Etienne. 2012. ⟨hal-00714483⟩