Method

Densely Constrained Depth Estimator for Monocular 3D Object Detection [DCD]
-

Submitted on 5 Mar. 2022 08:54 by
Cong Pan (UCAS)

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

Method Description:
ECCV2022: https://arxiv.org/abs/2207.10047
Parameters:
-
Latex Bibtex:
@inproceedings{li2022densely,
title={Densely Constrained Depth Estimator for
Monocular 3D Object Detection},
author={Li, Yingyan and Chen, Yuntao and He,
Jiawei and Zhang, Zhaoxiang},
booktitle={European Conference on Computer
Vision},
pages={718--734},
year={2022},
organization={Springer}
}

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) 96.44 % 90.93 % 83.36 %
Car (Orientation) 96.31 % 90.66 % 83.01 %
Car (3D Detection) 23.81 % 15.90 % 13.21 %
Car (Bird's Eye View) 32.55 % 21.50 % 18.25 %
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