VIRD: View-Invariant Representation through Dual-Axis Transformation for Cross-View Pose Estimation

2026-03-13

AI Summary

This work addresses the challenge of sparse spatial correspondences and inaccurate pose estimation arising from the substantial viewpoint disparity between ground-level and satellite imagery. To bridge this cross-view gap, the authors propose a dual-axis transformation strategy: polar coordinate transformation establishes horizontal correspondences, while a context-enhanced positional attention mechanism aligns vertical structures. This explicit alignment is further reinforced by a novel view reconstruction loss that promotes both viewpoint invariance and cross-view consistency, enabling high-precision pose estimation without requiring prior orientation information. Evaluated on the KITTI and VIGOR benchmarks, the method reduces position errors by 50.7% and 76.5%, and orientation errors by 18.0% and 46.8%, respectively, significantly outperforming existing approaches.

Technology Category

• Computer Vision: Multi-modal Vision
• Intelligent Robots: State Estimation
• Machine Learning: Multi-instance/Multi-view Learning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Accurate global localization is crucial for autonomous driving and robotics, but GNSS-based approaches often degrade due to occlusion and multipath effects. As an emerging alternative, cross-view pose estimation predicts the 3-DoF camera pose corresponding to a ground-view image with respect to a geo-referenced satellite image. However, existing methods struggle to bridge the significant viewpoint gap between the ground and satellite views mainly due to limited spatial correspondences. We propose a novel cross-view pose estimation method that constructs view-invariant representations through dual-axis transformation (VIRD). VIRD first applies a polar transformation to the satellite view to establish horizontal correspondence, then uses context-enhanced positional attention on the ground and polar-transformed satellite features to resolve vertical misalignment, explicitly mitigating the viewpoint gap. A view-reconstruction loss is introduced to strengthen the view invariance further, encouraging the derived representations to reconstruct the original and cross-view images. Experiments on the KITTI and VIGOR datasets demonstrate that VIRD outperforms the state-of-the-art methods without orientation priors, reducing median position and orientation errors by 50.7% and 76.5% on KITTI, and 18.0% and 46.8% on VIGOR, respectively.

Problem

• cross-view pose estimation
• viewpoint gap
• spatial correspondences
• view-invariant representation
• camera pose

Innovation

• dual-axis transformation
• view-invariant representation
• cross-view pose estimation
• polar transformation
• positional attention