FocalFormer3D : Focusing on Hard Instance for 3D Object Detection. (arXiv:2308.04556v1 [cs.CV])

False negatives (FN) in 3D object detection, {\em e.g.}, missing predictions
of pedestrians, vehicles, or other obstacles, can lead to potentially dangerous
situations in autonomous driving. While being fatal, this issue is understudied
in many current 3D detection methods. In this work, we propose Hard Instance
Probing (HIP), a general pipeline that identifies \textit{FN} in a multi-stage
manner and guides the models to focus on excavating difficult instances. For 3D
object detection, we instantiate this method as FocalFormer3D, a simple …

3d object detection arxiv autonomous autonomous driving current detection detection methods driving false instance issue pedestrians predictions work