Eigenboosting: Combining Discriminative and Generative Information

Horst Bischof (TU Graz, Austria)

A major shortcoming of discriminative recognition and detection methods is their noise sensitivity, both during training and recognition. This may lead to very sensitive and brittle recognition systems focusing on irrelevant information. This paper proposes to augment a discriminative classifier by reconstructive (generative) information. In particular, we boost classical Haar-like features and use the same features to approximate a generative model (\ie, eigenimages). A modified error function for boosting ensures that only features are selected that show a good discrimination and reconstruction. This allows a robust feature selection using boosting. Thus, we can handle problems where discriminant classifiers usually fail while still retaining the discriminative power. Our experiments show that we can significantly improve the recognition performance when learning from noisy data. Moreover, the used feature type allows efficient recognition and reconstruction.