Learning Segmentation by Random Walks

Marina Meila, Jianbo Shi

Abstract: We present a new view of image segmentation by pairwise simi(cid:173) larities. We interpret the similarities as edge flows in a Markov random walk and study the eigenvalues and eigenvectors of the walk's transition matrix. This interpretation shows that spectral methods for clustering and segmentation have a probabilistic foun(cid:173) dation. In particular, we prove that the Normalized Cut method arises naturally from our framework. Finally, the framework pro(cid:173) vides a principled method for learning the similarity function as a combination of features.