Method

DFR-Net [DFR-Net]


Submitted on 16 Mar. 2021 03:59 by
Lucas (Fudan)

Running time:0.18 s
Environment:1080 Ti (Pytorch)

Method Description:
Low-cost monocular 3D object detection plays a
fundamental role in autonomous driving, whereas
its accuracy is still far from satisfactory. Our
objective is to dig into the 3D object detection
task and reformulate it as the sub-tasks of
object localization and appearance perception,
which benefits to a deep excavation of reciprocal
information underlying the entire task. We
introduce a Dynamic Feature Reflecting Network,
named DFR-Net, which contains two novel
standalone modules: (i) the Appearance-
Localization Feature Reflecting module (ALFR)
that first separates task-specific features and
then self-mutually reflects the reciprocal
features; (ii) the Dynamic Intra-Trading module
(DIT) that adaptively realigns the training
processes of various sub-tasks via a self-
learning manner. Extensive experiments on the
challenging KITTI dataset demonstrate the
effectiveness and generalization of DFR-Net. We
rank 1st among all the monocular 3D object
detectors in the KITTI test set (till March 16
Parameters:
TBD
Latex Bibtex:
@inproceedings{dfr2021,
title={
The devil is in the task: Exploiting reciprocal
appearance-localization features for monocular 3d
object detection
},
author={
Zou, Zhikang and Ye, Xiaoqing and Du, Liang and
Cheng, Xianhui and Tan, Xiao and Zhang, Li
and Feng, Jianfeng and Xue, Xiangyang and Ding,
Errui
},
booktitle={ICCV},
year={2021}
}

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) 90.13 % 78.81 % 60.40 %
Car (Orientation) 89.79 % 77.41 % 59.20 %
Car (3D Detection) 19.40 % 13.63 % 10.35 %
Car (Bird's Eye View) 28.17 % 19.17 % 14.84 %
Pedestrian (Detection) 45.20 % 31.84 % 27.94 %
Pedestrian (Orientation) 35.75 % 24.88 % 21.72 %
Pedestrian (3D Detection) 6.09 % 3.62 % 3.39 %
Pedestrian (Bird's Eye View) 6.66 % 4.52 % 3.71 %
Cyclist (Detection) 48.34 % 31.93 % 27.95 %
Cyclist (Orientation) 38.60 % 24.85 % 21.86 %
Cyclist (3D Detection) 5.69 % 3.58 % 3.10 %
Cyclist (Bird's Eye View) 5.99 % 4.00 % 3.95 %
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



3D object detection results.
This figure as: png eps pdf txt gnuplot



Bird's eye view 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



3D object detection results.
This figure as: png eps pdf txt gnuplot



Bird's eye view 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



3D object detection results.
This figure as: png eps pdf txt gnuplot



Bird's eye view results.
This figure as: png eps pdf txt gnuplot




eXTReMe Tracker