FreeReg: Image-to-Point Cloud Registration Leveraging Pretrained Diffusion Models and Monocular Depth Estimators

TL;DR

This paper introduces FreeReg, a novel image-to-point cloud registration method that leverages pretrained diffusion models and monocular depth estimators to establish robust cross-modality correspondences without task-specific training.

Contribution

The paper proposes using diffusion features from pretrained models and geometric features from depth maps to improve cross-modality registration accuracy without additional training.

Findings

Achieves 20.6% higher Inlier Ratio on benchmarks.

Triples the Inlier Number compared to previous methods.

Improves registration recall by 48.6% over state-of-the-art.

Abstract

Matching cross-modality features between images and point clouds is a fundamental problem for image-to-point cloud registration. However, due to the modality difference between images and points, it is difficult to learn robust and discriminative cross-modality features by existing metric learning methods for feature matching. Instead of applying metric learning on cross-modality data, we propose to unify the modality between images and point clouds by pretrained large-scale models first, and then establish robust correspondence within the same modality. We show that the intermediate features, called diffusion features, extracted by depth-to-image diffusion models are semantically consistent between images and point clouds, which enables the building of coarse but robust cross-modality correspondences. We further extract geometric features on depth maps produced by the monocular depth estimator. By matching such geometric features, we significantly improve the accuracy of the coarse correspondences produced by diffusion features. Extensive experiments demonstrate that without any task-specific training, direct utilization of both features produces accurate image-to-point cloud registration. On three public indoor and outdoor benchmarks, the proposed method averagely achieves a 20.6 percent improvement in Inlier Ratio, a three-fold higher Inlier Number, and a 48.6 percent improvement in Registration Recall than existing state-of-the-arts.