Universe Points Representation Learning for Partial Multi-Graph Matching

TL;DR

This paper introduces URL, a novel deep learning method for partial multi-graph matching that leverages universe points and latent representations, improving scalability and robustness in semantic keypoint matching tasks.

Contribution

The paper presents a new universe points-based deep learning approach for partial multi-graph matching with cycle consistency, advancing the state of the art.

Findings

Outperforms existing methods on Pascal VOC, CUB, and Willow datasets.

Demonstrates scalability to large graphs with many nodes.

Shows robustness to high levels of partiality in synthetic experiments.

Abstract

Many challenges from natural world can be formulated as a graph matching problem. Previous deep learning-based methods mainly consider a full two-graph matching setting. In this work, we study the more general partial matching problem with multi-graph cycle consistency guarantees. Building on a recent progress in deep learning on graphs, we propose a novel data-driven method (URL) for partial multi-graph matching, which uses an object-to-universe formulation and learns latent representations of abstract universe points. The proposed approach advances the state of the art in semantic keypoint matching problem, evaluated on Pascal VOC, CUB, and Willow datasets. Moreover, the set of controlled experiments on a synthetic graph matching dataset demonstrates the scalability of our method to graphs with large number of nodes and its robustness to high partiality.