RoboFusion: Towards Robust Multi-Modal 3D Object Detection via SAM

arXiv:2401.03907v2 Announce Type: replace
Abstract: Multi-modal 3D object detectors are dedicated to exploring secure and reliable perception systems for autonomous driving (AD). However, while achieving state-of-the-art (SOTA) performance on clean benchmark datasets, they tend to overlook the complexity and harsh conditions of real-world environments. Meanwhile, with the emergence of visual foundation models (VFMs), opportunities and challenges are presented for improving the robustness and generalization of multi-modal 3D object detection in autonomous driving. Therefore, we propose RoboFusion, a robust framework that …

3d object 3d object detection abstract art arxiv autonomous autonomous driving benchmark complexity cs.cv datasets detection detectors driving emergence environments foundation however modal multi-modal object perception performance robust sam sota state systems type via visual world