Learning Propagation for Arbitrarily-structured Data

TL;DR

This paper introduces Spatial Generalized Propagation Networks (SGPN), a learnable diffusion-based module that enhances semantic segmentation across various arbitrarily-structured data types like images, superpixels, and point clouds.

Contribution

The paper proposes a novel, flexible propagation module that models global pairwise relations in arbitrarily-structured data, improving segmentation performance.

Findings

SGPN improves segmentation accuracy on images, superpixels, and point clouds.

The module is compatible with various neural network architectures.

Experiments show consistent performance gains over baseline models.

Abstract

Processing an input signal that contains arbitrary structures, e.g., superpixels and point clouds, remains a big challenge in computer vision. Linear diffusion, an effective model for image processing, has been recently integrated with deep learning algorithms. In this paper, we propose to learn pairwise relations among data points in a global fashion to improve semantic segmentation with arbitrarily-structured data, through spatial generalized propagation networks (SGPN). The network propagates information on a group of graphs, which represent the arbitrarily-structured data, through a learned, linear diffusion process. The module is flexible to be embedded and jointly trained with many types of networks, e.g., CNNs. We experiment with semantic segmentation networks, where we use our propagation module to jointly train on different data -- images, superpixels and point clouds. We show that SGPN consistently improves the performance of both pixel and point cloud segmentation, compared to networks that do not contain this module. Our method suggests an effective way to model the global pairwise relations for arbitrarily-structured data.