Abstract
Sparse subspace clustering(SSC).as one of the most successfiuul subspace clustering mcthods,has achieved no-table clustering accuracy in computer vision tasks.Howev-er,SSC applies only to vector data in Euclidean space.Un-fortunately there is still no satisfactory approach to solve subspace clustering by self -expressive principle for sym-metric positive definite(SPD)matrices which is very useful in computer vision.In this paper by embedding the SPD matrices into a Rcproducing Kernel Hilbcrt Space(RKHS),a kernel subspace clustering method is constructed on the SPD manifold through an appropriate Log-Euclidean ker-nel,termed as kernel sparse subspace clustering on the SPD Riemannian manifold(KSSCR).By exploiting the intrinsic Ricmannian gcomctry within data,KSSCR can effcctively characterize the geodesic distance between SPD matrices to uncover the underlying subspace strucfure.Erperimen-tul resulis on severul fumous datasets demonstrute that the proposed method achieves better clustering results than the state-of-the-art approaches.