Deep Generic Dynamic Object Detection Based on Dynamic Grid Maps

TL;DR

This paper introduces a deep learning method using dynamic grid maps for detecting various dynamic objects in automated driving, significantly reducing false positives compared to traditional clustering methods.

Contribution

It presents a novel approach combining LiDAR-based dynamic grid generation with a rotation-equivariant deep detector for improved object detection in autonomous vehicles.

Findings

Reduced false positive detection rate

Effective detection of arbitrary dynamic objects

Outperforms classic clustering strategies

Abstract

This paper describes a method to detect generic dynamic objects for automated driving. First, a LiDAR-based dynamic grid is generated online. Second, a deep learning-based detector is trained on the dynamic grid to infer the presence of dynamic objects of any type, which is a prerequisite for safe automated vehicles in arbitrary, edge-case scenarios. The Rotation-equivariant Detector (ReDet) - originally designed for oriented object detection on aerial images - was chosen due to its high detection performance. Experiments are conducted based on real sensor data and the benefits in comparison to classic dynamic cell clustering strategies are highlighted. The false positive object detection rate is strongly reduced by the proposed approach.