Method

Fusion_Net_V2 [FNV2]
[Anonymous Submission]

Submitted on 2 Aug. 2018 07:29 by
[Anonymous Submission]

Running time:0.18 s
Environment:GPU @ 2.5 Ghz (Python)

Method Description:
Fusion Net V2,[Fusion Front PC and RGB], train.txt,
First Try. From Deep_Earth_GO....
Parameters:
Not Published
Latex Bibtex:

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) 90.51 % 89.88 % 80.66 %
Car (3D Detection) 67.67 % 59.26 % 51.97 %
Car (Bird's Eye View) 80.68 % 74.36 % 65.88 %
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