A Graph-based Vehicle Proposal Location and Multi-View Classifiers Selection Algorithms [GVPL]

Submitted on 13 Apr. 2016 09:16 by
Su Shuai (Beijing Jiaotong Univercity,China)

Running time:1 s
Environment:8 cores @ 2.5 Ghz (Matlab + C/C++)

Method Description:
1.A Graph-based Vehicle Proposal Location
2.Multi-View Classifiers Selection Algorithms
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) 77.74 % 67.78 % 57.69 %
Car (Orientation) 61.33 % 54.32 % 46.24 %
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

eXTReMe Tracker