T-Graph: Enhancing Sparse-view Camera Pose Estimation by Pairwise Translation Graph

TL;DR

T-Graph is a novel module that leverages pairwise translation relationships to significantly improve sparse-view camera pose estimation, enhancing accuracy and robustness across different models and datasets.

Contribution

We propose T-Graph, a lightweight, plug-and-play translation graph module that incorporates pairwise translation information to boost sparse-view camera pose estimation performance.

Findings

Improves camera center accuracy by 1% to 6% across 2 to 8 viewpoints.

Validates effectiveness on state-of-the-art methods and public datasets.

Enhances robustness with two novel pairwise translation representations.

Abstract

Sparse-view camera pose estimation, which aims to estimate the 6-Degree-of-Freedom (6-DoF) poses from a limited number of images captured from different viewpoints, is a fundamental yet challenging problem in remote sensing applications. Existing methods often overlook the translation information between each pair of viewpoints, leading to suboptimal performance in sparse-view scenarios. To address this limitation, we introduce T-Graph, a lightweight, plug-and-play module to enhance camera pose estimation in sparse-view settings. T-graph takes paired image features as input and maps them through a Multilayer Perceptron (MLP). It then constructs a fully connected translation graph, where nodes represent cameras and edges encode their translation relationships. It can be seamlessly integrated into existing models as an additional branch in parallel with the original prediction, maintaining efficiency and ease of use. Furthermore, we introduce two pairwise translation representations, relative-t and pair-t, formulated under different local coordinate systems. While relative-t captures intuitive spatial relationships, pair-t offers a rotation-disentangled alternative. The two representations contribute to enhanced adaptability across diverse application scenarios, further improving our module's robustness. Extensive experiments on two state-of-the-art methods (RelPose++ and Forge) using public datasets (C03D and IMC PhotoTourism) validate both the effectiveness and generalizability of T-Graph. The results demonstrate consistent improvements across various metrics, notably camera center accuracy, which improves by 1% to 6% from 2 to 8 viewpoints.