KIP-Veröffentlichungen

Jahr 2011
Autor(en) J. Wagner, B. Ommer
Titel Efficient Clustering Earth Mover"s Distance
KIP-Nummer HD-KIP 11-25
KIP-Gruppe(n) F18
Dokumentart Paper
Abstract (en)

The two-class clustering problem is formulated as an integer convex optimisation problem which determines the maximum of the Earth Mover’s Distance (EMD) between two classes, constructing a bipartite graph with minimum flow and maximum inter-class EMD between two sets. Subsequently including the nearest neighbours of the start point in feature space and calculating the EMD for this labellings quickly converges to a robust optimum. A histogram of grey values with the number of bins $b$ as the only parameter is used as feature, which makes run time complexity independent of the number of pixels. After convergence in $\mathcal{O}(b)$ steps, spatial correlations can be taken into account by total variational smoothing. Testing the algorithm on real world images from commonly used databases reveals that it is competitive to state-of-the-art methods, while it deterministically yields hard assignments without requiring any a priori knowledge of the input data or similarity matrices to be calculated.

KIP - Bibliothek
Im Neuenheimer Feld 227
Raum 3.402
69120 Heidelberg