Method

DID-M3D [DID-M3D]
https://github.com/SPengLiang/DID-M3D

Submitted on 5 Mar. 2022 09:04 by
Xiaopei Wu (ZheJiang University)

Running time:0.04 s
Environment:1 core @ 2.5 Ghz (Python)

Method Description:
TBD
Parameters:
TBD
Latex Bibtex:
@inproceedings{peng2022did,
title={DID-M3D: Decoupling Instance Depth for
Monocular 3D Object Detection},
author={Peng, Liang and Wu, Xiaopei and Yang, Zheng
and Liu, Haifeng and Cai, Deng},
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) 94.29 % 91.04 % 81.31 %
Car (Orientation) 94.20 % 90.55 % 80.61 %
Car (3D Detection) 24.40 % 16.29 % 13.75 %
Car (Bird's Eye View) 32.95 % 22.76 % 19.83 %
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