Method

From Voxel to Point: IoU-guided 3D Object Detection for Point Cloud with Voxel-to-Point Decoder [FromVoxelToPoint]
https://github.com/jialeli1/From-Voxel-to-Point

Submitted on 19 Dec. 2020 10:04 by
Jiale Li (Zhejiang University)

Running time:0.1 s
Environment:1 core @ 2.5 Ghz (C/C++)

Method Description:
See paper.
Parameters:
See paper.
Latex Bibtex:
@inproceedings{FromVoxelToPoint,
author = {Jiale Li and
Hang Dai and
Ling Shao and
Yong Ding},
title = {From Voxel to Point: IoU-guided 3D
Object Detection for Point Cloud with Voxel-to-
Point Decoder},
booktitle = {{MM} '21: The 29th {ACM}
International Conference on Multimedia (ACM MM)},
publisher = {{ACM}},
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) 96.08 % 93.06 % 90.53 %
Car (Orientation) 96.07 % 92.98 % 90.40 %
Car (3D Detection) 88.53 % 81.58 % 77.37 %
Car (Bird's Eye View) 92.23 % 88.61 % 86.11 %
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