Method

Rethinking IoU-based Optimization for Single-stage 3D Object Detection [RDIoU]
https://github.com/hlsheng1/RDIoU

Submitted on 4 Jul. 2022 07:52 by
Hualian Sheng (Zhejiang University)

Running time:0.03 s
Environment:1 core @ 2.5 Ghz (Python + C/C++)

Method Description:
TBD
Parameters:
TBD
Latex Bibtex:
@inproceedings{sheng2022rdiou,
title={Rethinking IoU-based Optimization for Single-
stage 3D Object Detection},
author={Sheng, Hualian and Cai, Sijia and Zhao, Na
and Deng, Bing and Huang, Jianqiang and Hua, Xian-
Sheng and Zhao, Min-Jian and Lee, Hee, Gim},
booktitle={ECCV},
year={2022}
}

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) 98.79 % 96.05 % 91.03 %
Car (Orientation) 98.77 % 95.95 % 90.90 %
Car (3D Detection) 90.65 % 82.30 % 77.26 %
Car (Bird's Eye View) 94.90 % 89.75 % 84.67 %
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