Method

3D Bounding Box Estimation Using Deep Learning and Geometry [Deep3DBox]


Submitted on 1 Nov. 2016 18:32 by
Arsalan Mousavian (George Mason University)

Running time:1.5 s
Environment:GPU @ 2.5 Ghz (C/C++)

Method Description:
https://arxiv.org/abs/1612.00496
Parameters:
Latex Bibtex:
@inproceedings{MousavianCVPR2017,
title = {3D Bounding Box Estimation Using Deep
Learning and Geometry},
author = {Arsalan Mousavian and Dragomir Anguelov
and John Flynn and Jana Kosecka},
booktitle = {CVPR},
year = {2017}
}

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) 94.71 % 90.19 % 76.82 %
Car (Orientation) 94.62 % 89.88 % 76.40 %
Cyclist (Detection) 84.36 % 74.78 % 64.05 %
Cyclist (Orientation) 68.31 % 58.56 % 50.30 %
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



2D object detection results.
This figure as: png eps pdf txt gnuplot



Orientation estimation results.
This figure as: png eps pdf txt gnuplot




eXTReMe Tracker