Method

epBRM[la] [epBRM]
https://arxiv.org/abs/1910.04853

Submitted on 20 Apr. 2019 08:04 by
Kiwoo Shin (UC Berkeley)

Running time:0.10 s
Environment:1 core @ 2.5 Ghz (C/C++)

Method Description:
Improving a Quality of 3D Object Detection by
Spatial Transformation Mechanism

Please also read my previous paper:
RoarNet(https://arxiv.org/abs/1811.03818)
Parameters:
N/A
Latex Bibtex:
@article{arxiv,
title={Improving a Quality of 3D Object Detection
by Spatial Transformation Mechanism},
author={K Shin},
journal={arXiv preprint arXiv:1910.04853},
year={2019}
}

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
Pedestrian (Detection) 62.90 % 54.13 % 51.95 %
Pedestrian (3D Detection) 49.17 % 41.52 % 39.08 %
Pedestrian (Bird's Eye View) 52.48 % 45.49 % 42.75 %
This table as LaTeX

2D object detection 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