Iterative Camera-LiDAR Extrinsic Optimization via Surrogate Diffusion

2026-04-11

AI Summary

To address the trade-off between accuracy and efficiency in camera–LiDAR extrinsic calibration for autonomous driving, this paper proposes a single-model iterative optimization method based on surrogate diffusion. Our key contributions are: (1) the first surrogate diffusion framework, eliminating training overhead and storage burden associated with multi-model ensembles; (2) a dual-path denoising network—projection-first and encoding-first—that jointly enhances point-cloud projection consistency and feature representation capability; and (3) a buffered inference mechanism reducing inference latency by 43.7%. Experiments demonstrate that our method achieves a 24.5% reduction in rotational error over the second-best approach, and further improves upon a baseline diffusion method by 20.4% (rotation) and 9.6% (translation). Its accuracy matches that of multi-range models while significantly outperforming existing single-model iterative methods.

Technology Category

Computer Vision: Diffusion Models for VisionSearch and Optimization: Learning to SearchMachine Learning: Calibration & Uncertainty Quantification

Application Category

Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMsSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for rankingUser Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Cameras and LiDAR are essential sensors for autonomous vehicles. Camera-LiDAR data fusion compensate for deficiencies of stand-alone sensors but relies on precise extrinsic calibration. Many learning-based calibration methods predict extrinsic parameters in a single step. Driven by the growing demand for higher accuracy, a few approaches utilize multi-range models or integrate multiple methods to improve extrinsic parameter predictions, but these strategies incur extended training times and require additional storage for separate models. To address these issues, we propose a single-model iterative approach based on surrogate diffusion to significantly enhance the capacity of individual calibration methods. By applying a buffering technique proposed by us, the inference time of our surrogate diffusion is 43.7% less than that of multi-range models. Additionally, we create a calibration network as our denoiser, featuring both projection-first and encoding-first branches for effective point feature extraction. Extensive experiments demonstrate that our diffusion model outperforms other single-model iterative methods and delivers competitive results compared to multi-range models. Our denoiser exceeds state-of-the-art calibration methods, reducing the rotation error by 24.5% compared to the second-best method. Furthermore, with the proposed diffusion applied, it achieves 20.4% less rotation error and 9.6% less translation error.

Problem

• Enhancing Camera-LiDAR extrinsic calibration accuracy
• Reducing training time and model storage requirements
• Improving point feature extraction for better fusion

Innovation

• Single-model iterative approach via surrogate diffusion
• Buffering technique reduces inference time by 43.7%
• Dual-branch denoiser for effective point feature extraction