Method

guided_anchor_50_2500_full_data [GA_FULLDATA]


Submitted on 19 Dec. 2019 14:56 by
Zhuohui Zhang (Hefei University of Technology)

Running time:1 s
Environment:4 cores @ 2.5 Ghz (Python)

Method Description:
guided_anchor_50_2500_full_data
Parameters:
20epochs
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) 96.31 % 90.73 % 78.22 %
Car (Orientation) 38.90 % 36.43 % 31.61 %
Cyclist (Detection) 75.50 % 57.20 % 50.26 %
Cyclist (Orientation) 33.35 % 25.80 % 22.70 %
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