Method

Huawei_Media_SSD_RCNN [HM_SSD_RCNN]


Submitted on 13 Dec. 2016 03:07 by
Wei Zhang (SIAT)

Running time:0.15 s
Environment:1 core @ 2.5 Ghz (C/C++)

Method Description:
1.SSD 512 for robust object proposal extraction
2.Finetune object detection with RCNN
Parameters:
Default setting for SSD & Faster-RCNN
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) 90.47 % 88.69 % 77.86 %
Car (Orientation) 46.40 % 44.44 % 39.98 %
Pedestrian (Detection) 82.33 % 66.41 % 59.21 %
Pedestrian (Orientation) 42.11 % 34.38 % 30.73 %
Cyclist (Detection) 77.55 % 64.67 % 54.70 %
Cyclist (Orientation) 44.70 % 35.64 % 29.65 %
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



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