Method

Learning a Layout Transfer Network for Context Aware Object Detection [LTN]


Submitted on 15 Sep. 2019 10:54 by
Tao Wang (Minjiang University)

Running time:0.4 s
Environment:GPU @ >3.5 Ghz (Python)

Method Description:
We present a context aware object detection
method based on a retrieve-and-transform scene
layout model. Given an input image, our approach
first retrieves a coarse scene layout from a
codebook of typical layout templates. In order to
handle large layout variations, we use a variant
of the spatial transformer network to transform
and refine the retrieved layout, resulting in a
set of interpretable and semantically meaningful
feature maps of object locations and scales. The
above steps are implemented as a Layout Transfer
Network which we integrate into Faster RCNN to
allow for joint reasoning of object detection and
scene layout estimation. Extensive experiments on
three public datasets verified that our approach
provides consistent performance improvements to
the state-of-the-art object detection baselines
on a variety of challenging tasks in the traffic
surveillance and the autonomous driving domains.
Parameters:
Please see the paper.
Latex Bibtex:
@ARTICLE{8836625,
author={T. {Wang} and X. {He} and Y. {Cai} and G.
{Xiao}},
journal={IEEE Transactions on Intelligent
Transportation Systems},
title={Learning a Layout Transfer Network for
Context Aware Object Detection},
year={2019},
volume={},
number={},
pages={1-16},
keywords={Object detection;context modeling;scene
layout transfer.},
doi={10.1109/TITS.2019.2939213},
ISSN={},
month={},}

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) 94.68 % 91.18 % 81.51 %
Car (Orientation) 48.96 % 46.54 % 41.58 %
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