Multi-Order Matching Network for Alignment-Free Depth Super-Resolution

TL;DR

MOMNet is an alignment-free depth super-resolution framework that adaptively retrieves relevant RGB information from misaligned scenes using multi-order matching and aggregation, outperforming existing methods.

Contribution

The paper introduces MOMNet, a novel alignment-free approach utilizing multi-order matching and aggregation to improve depth super-resolution in misaligned RGB-D data.

Findings

MOMNet outperforms existing methods on unaligned and aligned datasets.

The multi-order matching mechanism effectively identifies relevant RGB features.

Multi-order aggregation enhances feature transfer from RGB to depth.

Abstract

Recent guided depth super-resolution methods are premised on the assumption of strict spatial alignment between depth and RGB, achieving high-quality depth reconstruction. However, in real-world scenarios, the acquisition of strictly aligned RGB-D is hindered by inherent hardware limitations (e.g., physically separate RGB-D sensors) and unavoidable calibration drift induced by mechanical vibrations or temperature variations. Consequently, existing approaches often suffer inevitable performance degradation when applied to misaligned real-world scenes. In this paper, we propose the Multi-Order Matching Network (MOMNet), a novel alignment-free framework that adaptively retrieves and selects the most relevant information from misaligned RGB. Specifically, our method begins with a multi-order matching mechanism, which jointly performs zero-order, first-order, and second-order matching to comprehensively identify RGB information consistent with depth across multi-order feature spaces. To effectively integrate the retrieved RGB and depth, we further introduce a multi-order aggregation composed of multiple structure detectors. This strategy uses multi-order priors as prompts to facilitate the selective feature transfer from RGB to depth. Extensive experiments demonstrate that MOMNet achieves superior performance and generalization across both unaligned and aligned datasets.