Method

HNet [HNet]
[Anonymous Submission]

Submitted on 6 Dec. 2017 10:12 by
[Anonymous Submission]

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

Method Description:
Enhanced faster rcnn
Parameters:
version = 1.0
Latex Bibtex:
null

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) 77.09 % 66.00 % 53.89 %
Pedestrian (Detection) 77.39 % 66.74 % 62.26 %
Cyclist (Detection) 69.71 % 54.10 % 48.02 %
This table as LaTeX


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



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



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




eXTReMe Tracker