Multi-Image Semantic Matching by Mining Consistent Features

TL;DR

This paper introduces a scalable multi-image semantic matching approach that identifies reliable features and enforces geometric consistency, enabling effective reconstruction and landmark discovery without annotations.

Contribution

It presents a novel sparse feature matching method with a low-rank constraint for geometric consistency across many images, improving scalability and robustness.

Findings

Achieves competitive results on multi-graph matching benchmarks.

Successfully reconstructs object-class models without annotations.

Discovers object landmarks from images without supervision.

Abstract

This work proposes a multi-image matching method to estimate semantic correspondences across multiple images. In contrast to the previous methods that optimize all pairwise correspondences, the proposed method identifies and matches only a sparse set of reliable features in the image collection. In this way, the proposed method is able to prune nonrepeatable features and also highly scalable to handle thousands of images. We additionally propose a low-rank constraint to ensure the geometric consistency of feature correspondences over the whole image collection. Besides the competitive performance on multi-graph matching and semantic flow benchmarks, we also demonstrate the applicability of the proposed method for reconstructing object-class models and discovering object-class landmarks from images without using any annotation.