IIS at ImageCLEF 2015: Multi-label Classification Task.
Antonio Jose Rodríguez-Sánchez, Sabrina Fontanella, Justus Piater, Sándor Szedmák · CLEF (Working Notes) · 2015
We propose an image decomposition technique that captures the structure of a scene. An image is decomposed into a matrix that represents the adjacency between the elements of the image and their distance. Images decomposed this way are then classied using a max- imum margin regression (MMR) approach where the normal vector of the separating hyperplane maps the input feature vectors into the out- puts vectors. Multiclass and multilabel classication are native to MMR, unlike other more classical maximum margin approaches, like SVM. We have tested our approach with the ImageCLEF 2015 multi-label classi- cation task, obtaining high rankings at that task.