Abstract. This contribution proposes a compositional approach to visual object categorization of scenes. Compositions are learned from the Caltech 101 database1 intermediate abstra...
Several state-of-the-art Generic Visual Categorization (GVC) systems are built around a vocabulary of visual terms and characterize images with one histogram of visual word counts....
Florent Perronnin, Christopher R. Dance, Gabriela ...
Abstract. Principal Component Analysis (PCA) is one of the most popular techniques for dimensionality reduction of multivariate data points with application areas covering many bra...
Abstract. Deformable models are by their formulation able to solve surface extraction problem from noisy volumetric images. This is since they use image independent information, in...
In recent years several techniques have been proposed for modelling the low-dimensional manifolds, or `subspaces', of natural images. Examples include principal component anal...