Radar+RGB Attentive Fusion for Robust Object Detection in Autonomous Vehicles

31 Aug 2020  ·  Ritu Yadav, Axel Vierling, Karsten Berns ·

This paper presents two variations of architecture referred to as RANet and BIRANet. The proposed architecture aims to use radar signal data along with RGB camera images to form a robust detection network that works efficiently, even in variable lighting and weather conditions such as rain, dust, fog, and others. First, radar information is fused in the feature extractor network. Second, radar points are used to generate guided anchors. Third, a method is proposed to improve region proposal network targets. BIRANet yields 72.3/75.3% average AP/AR on the NuScenes dataset, which is better than the performance of our base network Faster-RCNN with Feature pyramid network(FFPN). RANet gives 69.6/71.9% average AP/AR on the same dataset, which is reasonably acceptable performance. Also, both BIRANet and RANet are evaluated to be robust towards the noise.

PDF Abstract

Datasets


Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Object Detection nuScenes BIRANet(RGB+Radar) AP50 88.9 # 1
AP75 84.3 # 1
AR 75.3 # 1
MAP 72.3 # 1
AP85 65.7 # 1
AP(s) 53.5 # 1
AP(m) 70.1 # 1
AP(l) 76.9 # 1
AR(s) 56.2 # 1
AR(m) 73.2 # 1
AR(l) 79.8 # 1
Object Detection nuScenes RANet(Radar) AP50 83.9 # 2
AP75 80.1 # 2
AR 71.9 # 2
MAP 69 # 2
AP85 64.4 # 2
AP(s) 44.8 # 2
AP(m) 67.8 # 2
AP(l) 73.3 # 2
AR(s) 47.3 # 2
AR(m) 70.9 # 2
AR(l) 76.2 # 2

Methods


No methods listed for this paper. Add relevant methods here