Method

YOLOv3 + distance [YOLOv3+d]
[Anonymous Submission]

Submitted on 30 Jul. 2018 03:43 by
[Anonymous Submission]

Running time:0.04 s
Environment:GPU @ 1.5 Ghz (C/C++)

Method Description:
YOLOv3 trained with distance
Parameters:
width=1248
height=384
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) 84.30 % 84.13 % 76.34 %
Pedestrian (Detection) 67.23 % 51.03 % 48.87 %
Cyclist (Detection) 59.08 % 42.60 % 40.77 %
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