Graph-CoVis: GNN-based Multi-view Panorama Global Pose Estimation
This addresses camera pose estimation for 360° panoramas in real-world indoor environments, presenting an incremental extension from two-view to multi-view learning.
The paper tackles the problem of wide-baseline camera pose estimation from multiple 360° panoramas by introducing Graph-CoVis, a Graph Neural Network that jointly learns co-visible structure and global motion, and shows competitive performance to state-of-the-art approaches on the ZInD dataset.
In this paper, we address the problem of wide-baseline camera pose estimation from a group of 360$^\circ$ panoramas under upright-camera assumption. Recent work has demonstrated the merit of deep-learning for end-to-end direct relative pose regression in 360$^\circ$ panorama pairs [11]. To exploit the benefits of multi-view logic in a learning-based framework, we introduce Graph-CoVis, which non-trivially extends CoVisPose [11] from relative two-view to global multi-view spherical camera pose estimation. Graph-CoVis is a novel Graph Neural Network based architecture that jointly learns the co-visible structure and global motion in an end-to-end and fully-supervised approach. Using the ZInD [4] dataset, which features real homes presenting wide-baselines, occlusion, and limited visual overlap, we show that our model performs competitively to state-of-the-art approaches.