Context-guided diffusion for label propagation on graphs

TL;DR

This paper introduces anisotropic diffusion on graphs using diffusivity tensors, leading to a new label propagation algorithm that enhances semi-supervised learning performance.

Contribution

It develops a novel anisotropic diffusion framework on graphs by discretizing diffusivity operators inspired by Riemannian manifolds, improving label propagation methods.

Findings

Significant performance improvements over existing diffusion algorithms.

Robust diffusivity operators enhance semi-supervised learning.

Framework applicable to various graph-based learning tasks.

Abstract

Existing approaches for diffusion on graphs, e.g., for label propagation, are mainly focused on isotropic diffusion, which is induced by the commonly-used graph Laplacian regularizer. Inspired by the success of diffusivity tensors for anisotropic diffusion in image processing, we presents anisotropic diffusion on graphs and the corresponding label propagation algorithm. We develop positive definite diffusivity operators on the vector bundles of Riemannian manifolds, and discretize them to diffusivity operators on graphs. This enables us to easily define new robust diffusivity operators which significantly improve semi-supervised learning performance over existing diffusion algorithms.