Method

Voxel Feature Pyramid Network [HRI-VoxelFPN]


Submitted on 20 Jun. 2019 13:29 by
Yihong Guan (Case Western Reserve University)

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

Method Description:
One stage 3D detector based on VoxelNet
Multi-scale voxelized feature aggregation
Parameters:
TBA
Latex Bibtex:
Latex Bibtex:
@article{Kuang2020voxelFPN,
title={{Voxel-FPN:}multi-scale voxel feature
aggregation in 3D object detection from point
clouds},
author={Kuang, Hongwu and Wang, Bei and An,
Jianping and Zhang, Ming and Zhang, Zehan},
journal = {sensors},
year = {2020},
url = {https://www.mdpi.com/1424-8220/20/3/704}

}

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) 96.65 % 91.44 % 86.18 %
Car (Orientation) 96.35 % 90.76 % 85.37 %
Car (3D Detection) 85.64 % 76.70 % 69.44 %
Car (Bird's Eye View) 92.75 % 87.21 % 79.82 %
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




eXTReMe Tracker