Mining Exemplars for Object Modelling Using Affinity Propagation

Shengping Xia, Jianjun Liu, Edwin R. Hancock

Abstract: This paper focusses on the problem of locating object class exemplars from a large corpus of images using affinity propagation. We use attributed relational graphs to represent groups of local invariant features together with their spatial arrangement. Rather than mining exemplars from the entire graph corpus, we prefer to cluster object specific exemplars. Firstly, we obtain an object specific cluster of graphs using similarity propagation. The popular affinity propagation method is then individually applied to each object specific cluster. Using this clustering method, we can obtain object specific exemplars together with a high precision for the data associated with each exemplar. Experiments are performed on over 80 K images spanning ~500 objects, and demonstrate the performance of the method in terms of efficiency, scalability.