L4DR: LiDAR-4DRadar Fusion for Weather-Robust 3D Object Detection

Xun Huang,Ziyu Xu,Hai Wu,Jinlong Wang,Qiming Xia,Yan Xia,Jonathan Li,Kyle Gao,Chenglu Wen,Cheng Wang

2024-08-30

Abstract:LiDAR-based vision systems are integral for 3D object detection, which is crucial for autonomous navigation. However, they suffer from performance degradation in adverse weather conditions due to the quality deterioration of LiDAR point clouds. Fusing LiDAR with the weather-robust 4D radar sensor is expected to solve this problem. However, the fusion of LiDAR and 4D radar is challenging because they differ significantly in terms of data quality and the degree of degradation in adverse weather. To address these issues, we introduce L4DR, a weather-robust 3D object detection method that effectively achieves LiDAR and 4D Radar fusion. Our L4DR includes Multi-Modal Encoding (MME) and Foreground-Aware Denoising (FAD) technique to reconcile sensor gaps, which is the first exploration of the complementarity of early fusion between LiDAR and 4D radar. Additionally, we design an Inter-Modal and Intra-Modal ({IM}2 ) parallel feature extraction backbone coupled with a Multi-Scale Gated Fusion (MSGF) module to counteract the varying degrees of sensor degradation under adverse weather conditions. Experimental evaluation on a VoD dataset with simulated fog proves that L4DR is more adaptable to changing weather conditions. It delivers a significant performance increase under different fog levels, improving the 3D mAP by up to 20.0% over the traditional LiDAR-only approach. Moreover, the results on the K-Radar dataset validate the consistent performance improvement of L4DR in real-world adverse weather conditions.

Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

The paper aims to address the issue of performance degradation in LiDAR-based 3D object detection under adverse weather conditions in autonomous driving. Specifically:

1. **Problem Description**: Under adverse weather conditions (such as fog, rain, snow, etc.), the quality of LiDAR point clouds significantly decreases, leading to reduced 3D object detection performance.
2. **Solution**: Improve detection performance by integrating 4D radar sensors, which have higher weather robustness. However, due to the significant differences in data quality between LiDAR and 4D radar, and the varying degrees of sensor degradation under adverse weather conditions, a new fusion method is required.
3. **Proposed Method**: The paper proposes the L4DR framework, which includes:
- Multi-Modal Encoder (MME): Achieves complementary fusion of LiDAR and 4D radar data during the encoding stage.
- Foreground-Aware Denoising (FAD): Filters out noise points in the 4D radar.
- Inter-Modal and Intra-Modal backbone ({IM}²): Combines Multi-Scale Gated Fusion (MSGF) modules to adapt to sensor degradation under different weather conditions.

Experimental results show that L4DR exhibits significant performance improvements under various adverse weather conditions (such as fog, rain, snow, etc.), particularly with a 20.0% increase in 3D mean Average Precision (mAP) in simulated fog environments. Additionally, its performance advantages are also validated under real adverse weather conditions.