Robust Near-Isometric Matching via Structured Learning of Graphical Models

TL;DR

This paper presents a structured graphical model for near-rigid shape matching that learns parameters to handle noise, appearance, and scale variations, resulting in improved robustness and accuracy.

Contribution

It introduces a novel graphical model that jointly learns appearance, distance, and angle features for robust near-isometric shape matching.

Findings

Significant improvement over recent models in accuracy.

Maintains similar computational efficiency.

Effectively captures scale and appearance variations.

Abstract

Models for near-rigid shape matching are typically based on distance-related features, in order to infer matches that are consistent with the isometric assumption. However, real shapes from image datasets, even when expected to be related by "almost isometric" transformations, are actually subject not only to noise but also, to some limited degree, to variations in appearance and scale. In this paper, we introduce a graphical model that parameterises appearance, distance, and angle features and we learn all of the involved parameters via structured prediction. The outcome is a model for near-rigid shape matching which is robust in the sense that it is able to capture the possibly limited but still important scale and appearance variations. Our experimental results reveal substantial improvements upon recent successful models, while maintaining similar running times.