RobustCalib: Robust Lidar-Camera Extrinsic Calibration with Consistency Learning

TL;DR

RobustCalib introduces a novel, automatic, single-shot LiDAR-camera extrinsic calibration method leveraging consistency learning, improving robustness and adaptability without requiring offline targets or iterative refinement.

Contribution

It proposes a new consistency learning framework with appearance and geometric losses for robust, automatic, and single-shot extrinsic calibration of LiDAR and camera systems.

Findings

Achieves accurate and robust calibration across various datasets.

Outperforms traditional and learning-based methods in robustness.

Enables efficient inference without iterative optimization.

Abstract

Current traditional methods for LiDAR-camera extrinsics estimation depend on offline targets and human efforts, while learning-based approaches resort to iterative refinement for calibration results, posing constraints on their generalization and application in on-board systems. In this paper, we propose a novel approach to address the extrinsic calibration problem in a robust, automatic, and single-shot manner. Instead of directly optimizing extrinsics, we leverage the consistency learning between LiDAR and camera to implement implicit re-calibartion. Specially, we introduce an appearance-consistency loss and a geometric-consistency loss to minimizing the inconsitency between the attrbutes (e.g., intensity and depth) of projected LiDAR points and the predicted ones. This design not only enhances adaptability to various scenarios but also enables a simple and efficient formulation during inference. We conduct comprehensive experiments on different datasets, and the results demonstrate that our method achieves accurate and robust performance. To promote further research and development in this area, we will release our model and code.