Cross-View Yaw Estimation in Location Uncertainty with Line-Aligning Yaw Scoring
For cross-view localization tasks (e.g., autonomous driving), this method provides a robust yaw estimation that disentangles yaw from translation, improving downstream 3-DoF localization.
The paper tackles cross-view yaw estimation under location uncertainty, achieving sub-degree yaw precision via a radially invariant line-consensus voting method (LAYS) that eliminates the need for accurate location. Experiments show significant gains, e.g., +28-45%p for normal FoV with unknown yaw.
Accurate yaw estimation is a bottleneck in cross-view localization between ground view and Bird's Eye View (BEV). Existing methods couple yaw with translation and rely on height or projection assumptions that degrade under large yaw ambiguity. We disentangle yaw from location accuracy and introduce LAYS, a radially invariant line-consensus voting method. By exploiting the radial invariance of our formulation, we achieve sub-degree yaw precision via 3D voting over all candidate poses, while eliminating the need for accurate location. Our key observation is that a ground-image column matched to BEV pixels induces the same yaw across all camera positions along the radial direction of the pixels. LAYS matches BEV pixels to ground columns using feature similarity and accumulates the induced yaw votes into discrete 3D bins, where correct correspondences along the radial line concentrate into a sharp peak for the correct yaw. Experiments on Mapillary, Ford, KITTI, and VIGOR show significant gains under unknown yaw, particularly for normal FoV with unknown yaw (+28$\sim$45\%p), and using LAYS as a yaw prior improves downstream 3-DoF localization.