Method

SMOKE [SMOKE]
https://github.com/lzccccc/SMOKE

Submitted on 12 Nov. 2019 03:02 by
Zechen Liu (ZongMu)

Running time:0.03 s
Environment:GPU @ 2.5 Ghz (Python)

Method Description:
TBD
Parameters:
TBD
Latex Bibtex:
@misc{liu2020smoke,
title={SMOKE: Single-Stage Monocular 3D Object
Detection via Keypoint Estimation},
author={Zechen Liu and Zizhang Wu and Roland
Tóth},
year={2020},
eprint={2002.10111},
archivePrefix={arXiv},
primaryClass={cs.CV}
}

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) 93.21 % 87.51 % 77.66 %
Car (Orientation) 92.94 % 87.02 % 77.12 %
Car (3D Detection) 14.03 % 9.76 % 7.84 %
Car (Bird's Eye View) 20.83 % 14.49 % 12.75 %
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