Method

Geometry-based Distance Decomposition for Monocular 3D Object Detection [MonoRCNN]
https://github.com/Rock-100/MonoDet

Submitted on 27 Feb. 2021 03:41 by
Jack (Imperial College London)

Running time:0.07 s
Environment:GPU @ 2.5 Ghz (Python)

Method Description:
TBD
Parameters:
TBD
Latex Bibtex:
@inproceedings{MonoRCNN_ICCV21,
title = {Geometry-based Distance Decomposition for
Monocular 3D Object Detection},
author = {Xuepeng Shi and Qi Ye and
Xiaozhi Chen and Chuangrong Chen and
Zhixiang Chen and Tae-Kyun Kim},
booktitle = {ICCV},
year = {2021},
}

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) 91.98 % 86.78 % 66.97 %
Car (Orientation) 91.90 % 86.48 % 66.71 %
Car (3D Detection) 18.36 % 12.65 % 10.03 %
Car (Bird's Eye View) 25.48 % 18.11 % 14.10 %
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