LiREC-Net: A Target-Free and Learning-Based Network for LiDAR, RGB, and Event Calibration

TL;DR

LiREC-Net is a novel, target-free, learning-based network that jointly calibrates LiDAR, RGB, and event sensors for autonomous systems, improving multi-sensor fusion accuracy and efficiency.

Contribution

It introduces a unified framework for multi-modal sensor calibration, leveraging shared representations to enhance efficiency and performance in tri-modal calibration tasks.

Findings

Achieves competitive accuracy with bi-modal models

Sets a new baseline for tri-modal sensor calibration

Demonstrates effectiveness on KITTI and DSEC datasets

Abstract

Advanced autonomous systems rely on multi-sensor fusion for safer and more robust perception. To enable effective fusion, calibrating directly from natural driving scenes (i.e., target-free) with high accuracy is crucial for precise multi-sensor alignment. Existing learning-based calibration methods are typically designed for only a single pair of sensor modalities (i.e., a bi-modal setup). Unlike these methods, we propose LiREC-Net, a target-free, learning-based calibration network that jointly calibrates multiple sensor modality pairs, including LiDAR, RGB, and event data, within a unified framework. To reduce redundant computation and improve efficiency, we introduce a shared LiDAR representation that leverages features from both its 3D nature and projected depth map, ensuring better consistency across modalities. Trained and evaluated on established datasets, such as KITTI and DSEC, our LiREC-Net achieves competitive performance to bi-modal models and sets a new strong baseline for the tri-modal use case.