OCA: ODE-Driven Cross-Attention for Image-to-Point-Cloud Registration

2026-09-28

AI Summary

This study addresses the attention ambiguity problem inherent in cross-attention mechanisms for image–point cloud registration by proposing an ODE-driven cross-attention module. To our knowledge, this work is the first to introduce ordinary differential equations into cross-attention, modeling ideal feature interaction dynamics to jointly optimize the attention matrix and feature representations. This formulation significantly enhances the discriminability of 2D–3D feature correspondences and can be seamlessly integrated into existing frameworks. Extensive evaluations across four benchmark datasets demonstrate that the proposed method improves registration recall by 5%, 9%, and 15% under standard, fine-tuned, and zero-shot settings, respectively, thereby validating both its effectiveness and generalization capability.

Abstract

Cross-attention is a crucial component in learning-based image-to-point-cloud (I2P) registration. Although existing cross-attention mechanisms have achieved promising progress, attention ambiguity remains a fundamental challenge that hinders the learning of discriminative 2D-3D correspondences. To address this problem, we revisit cross-attention and establish ordinary differential equations (ODEs) to model the ideal I2P feature interaction. Based on this formulation, we develop an ODE-driven cross-attention (OCA) module that refines feature representations and attention matrices through ODEs. In practice, OCA can be seamlessly integrated into existing I2P registration frameworks. To validate its effectiveness, we incorporate OCA into five state-of-the-art baselines and evaluate on four public benchmark datasets. Experimental results demonstrate that OCA improves registration recall by up to 5\%, 9\%, and 15\% under the standard, fine-tuning, and zero-shot settings, respectively.

Problem

• Image-to-Point-Cloud Registration
• Cross-Attention
• Attention Ambiguity
• 2D-3D Correspondences

Innovation

• Image-to-Point-Cloud Registration
• Cross-Attention
• Ordinary Differential Equations (ODEs)
• Feature Interaction
• Zero-shot