Method

Baidu-RAL[la] [3D IoU Loss]


Submitted on 22 May. 2019 09:15 by
Dingfu ZHOU (BAIDU)

Running time:0.08 s
Environment:GPU @ 2.5 Ghz (Python + C/C++)

Method Description:
Rotated IoU loss is applied with 3D Object
Detection
Parameters:
Only lidar point cloud;
50/50 split for training;
The result is obtained with the SECOND method by
integrating the rotated IoU loss for 3D Object
detection.
Latex Bibtex:
@inproceedings{zhou2019,
title={IoU Loss for 2D/3D Object Detection},
author={Zhou, Dingfu and Fang, Jin and Song,
Xibin and Guan, Chenye and Yin, Junbo and Dai,
Yuchao and Yang, Ruigang},
booktitle={International Conference on 3D
Vision
(3DV)},
year={2019},
organization={IEEE}
}

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) 95.92 % 90.79 % 85.65 %
Car (Orientation) 95.60 % 90.21 % 84.96 %
Car (3D Detection) 86.16 % 76.50 % 71.39 %
Car (Bird's Eye View) 91.36 % 86.22 % 81.20 %
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