Sunshine to Rainstorm: Cross-Weather Knowledge Distillation for Robust 3D Object Detection

TL;DR

This paper introduces a novel rain simulation method and a cross-weather knowledge distillation approach to improve LiDAR-based 3D object detection robustness in rainy and sunny conditions, validated on large-scale datasets.

Contribution

It proposes DRET, a realistic rain simulation technique, and SRKD, a knowledge distillation framework, to enhance detection accuracy across diverse weather scenarios.

Findings

Significant accuracy improvements in rainy conditions.

Enhanced detection performance in sunny weather.

Compatibility with multiple 3D detection models.

Abstract

LiDAR-based 3D object detection models have traditionally struggled under rainy conditions due to the degraded and noisy scanning signals. Previous research has attempted to address this by simulating the noise from rain to improve the robustness of detection models. However, significant disparities exist between simulated and actual rain-impacted data points. In this work, we propose a novel rain simulation method, termed DRET, that unifies Dynamics and Rainy Environment Theory to provide a cost-effective means of expanding the available realistic rain data for 3D detection training. Furthermore, we present a Sunny-to-Rainy Knowledge Distillation (SRKD) approach to enhance 3D detection under rainy conditions. Extensive experiments on the WaymoOpenDataset large-scale dataset show that, when combined with the state-of-the-art DSVT model and other classical 3D detectors, our proposed framework demonstrates significant detection accuracy improvements, without losing efficiency. Remarkably, our framework also improves detection capabilities under sunny conditions, therefore offering a robust solution for 3D detection regardless of whether the weather is rainy or sunny