Relational Epipolar Graphs for Robust Relative Camera Pose Estimation
This work addresses the challenge of noisy correspondences in VSLAM for applications like robotics and augmented reality, representing an incremental improvement through a novel graph-based method.
The paper tackled the problem of robust relative camera pose estimation in VSLAM by reformulating it as a relational inference problem over epipolar correspondence graphs, achieving improved robustness to dense noise and large baseline variation on indoor and outdoor benchmarks compared to classical and learning-based methods.
A key component of Visual Simultaneous Localization and Mapping (VSLAM) is estimating relative camera poses using matched keypoints. Accurate estimation is challenged by noisy correspondences. Classical methods rely on stochastic hypothesis sampling and iterative estimation, while learning-based methods often lack explicit geometric structure. In this work, we reformulate relative pose estimation as a relational inference problem over epipolar correspondence graphs, where matched keypoints are nodes and nearby ones are connected by edges. Graph operations such as pruning, message passing, and pooling estimate a quaternion rotation, translation vector, and the Essential Matrix (EM). Minimizing a loss comprising (i) $\mathcal{L}_2$ differences with ground truth (GT), (ii) Frobenius norm between estimated and GT EMs, (iii) singular value differences, (iv) heading angle differences, and (v) scale differences, yields the relative pose between image pairs. The dense detector-free method LoFTR is used for matching. Experiments on indoor and outdoor benchmarks show improved robustness to dense noise and large baseline variation compared to classical and learning-guided approaches, highlighting the effectiveness of global relational consensus.