Method

RangeIoUDet: Range Image Based Real-Time 3D Object Detector Optimized by Intersection Over Union[la] [RangeIoUDet]


Submitted on 18 Nov. 2020 03:23 by
Zhidong Liang (Hikvision Research Institute)

Running time:0.02 s
Environment:GPU @ 2.5 Ghz (Python + C/C++)

Method Description:
See the paper.
Parameters:
See the paper.
Latex Bibtex:
@inproceedings{liang2021rangeioudet,
title={RangeIoUDet: Range Image Based Real-Time
3D
Object Detector Optimized by Intersection Over
Union},
author={Liang, Zhidong and Zhang, Zehan and
Zhang,
Ming and Zhao, Xian and Pu, Shiliang},
booktitle={CVPR},
pages={7140--7149},
year={2021}
}

Detailed Results

Object detection and orientation estimation results. Results for object detection are given in terms of average precision (AP) and results for joint object detection and orientation estimation are provided in terms of average orientation similarity (AOS).


Benchmark Easy Moderate Hard
Car (Detection) 95.74 % 94.61 % 91.98 %
Car (Orientation) 95.69 % 94.42 % 91.70 %
Car (3D Detection) 88.60 % 79.80 % 76.76 %
Car (Bird's Eye View) 92.28 % 88.59 % 85.83 %
Cyclist (Detection) 90.43 % 81.67 % 74.90 %
Cyclist (Orientation) 90.24 % 81.24 % 74.49 %
Cyclist (3D Detection) 83.12 % 67.77 % 60.26 %
Cyclist (Bird's Eye View) 85.99 % 71.49 % 63.62 %
This table as LaTeX


2D object detection results.
This figure as: png eps pdf txt gnuplot



Orientation estimation results.
This figure as: png eps pdf txt gnuplot



3D object detection results.
This figure as: png eps pdf txt gnuplot



Bird's eye view results.
This figure as: png eps pdf txt gnuplot



2D object detection results.
This figure as: png eps pdf txt gnuplot



Orientation estimation results.
This figure as: png eps pdf txt gnuplot



3D object detection results.
This figure as: png eps pdf txt gnuplot



Bird's eye view results.
This figure as: png eps pdf txt gnuplot




eXTReMe Tracker