End2End Multi-View Feature Matching with Differentiable Pose Optimization

Erroneous feature matches have severe impact on subsequent camera pose estimation and often require additional, time-costly measures, like RANSAC, for outlier rejection. Our method tackles this challenge by addressing feature matching and pose optimization jointly. To this end, we propose a graph attention network to predict image correspondences along with confidence weights. The resulting matches serve as weighted constraints in a differentiable pose estimation. Training feature matching with gradients from pose optimization naturally learns to down-weight outliers and boosts pose estimation on image pairs compared to SuperGlue by 6.7% on ScanNet. At the same time, it reduces the pose estimation time by over 50% and renders RANSAC iterations unnecessary. Moreover, we integrate information from multiple views by spanning the graph across multiple frames to predict the matches all at once. Multi-view matching combined with end-to-end training improves the pose estimation metrics on Matterport3D by 18.5% compared to SuperGlue.

End2End Multi-View Feature Matching connects feature matching and pose optimization in an end-to-end trainable approach that enables matches and confidences to be informed by the pose estimation objective. We introduce GNN-based multi-view matching to predict matches and confidences tailored to a differentiable pose solver, which significantly improves pose estimation performance.