Rolling Guidance Filter as a Clustering Algorithm

We propose a generalization of the rolling guidance filter (RGF) to similarity-based clustering (SBC) algorithm which can handle general vector data. The proposed RGF-based SBC makes similarities between data clearer than original similarity values computed from On basis values, we assign cluster labels by an algorithm. Experimental results show that achieves better result naive application values. Additionally, study convergence unimodal dataset its mean vector.