Method

RTM3D: Real-time Monocular 3D Detection from Object Keypoints for Autonomous Driving [RTM3D]
https://github.com/Banconxuan/RTM3D

Submitted on 26 Feb. 2020 08:49 by
peixuan li (University of Chinese Academy of Sciences)

Running time:0.05 s
Environment:GPU @ 1.0 Ghz (Python)

Method Description:
In this work, we propose an efficient and accurate
monocular 3D detection framework in single shot.
Most successful 3D detectors take the projection
constraint from the 3D bounding box to the 2D box
as an important component. Four edges of a 2D box
provide only four constraints and the performance
deteriorates dramatically with the small error of
the 2D detector. Different from these approaches,
our method predicts the nine perspective keypoints
of a 3D bounding box in image space, and then
utilize the geometric relationship of 3D and 2D
perspectives to recover the dimension, location,
and orientation in 3D space. In this method, the
properties of the object can be predicted stably
even when the estimation of keypoints is very
noisy, which enables us to obtain fast detection
speed with a small architecture. Training our
method only uses the 3D properties of the object
without the need for external networks or
supervision data. Our method is the first real-
time system for monocular image 3D detection while
achieves state-of-the-art performance on the KITTI
benchmark.
Parameters:
See paper for details.
Latex Bibtex:
@misc{li2020rtm3d,
title={RTM3D: Real-time Monocular 3D Detection
from Object Keypoints for Autonomous Driving},
author={Peixuan Li and Huaici Zhao and Pengfei
Liu and Feidao Cao},
year={2020},
eprint={2001.03343},
archivePrefix={arXiv},
primaryClass={cs.CV}
}

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) 91.82 % 86.93 % 77.41 %
Car (Orientation) 91.75 % 86.73 % 77.18 %
Car (3D Detection) 14.41 % 10.34 % 8.77 %
Car (Bird's Eye View) 19.17 % 14.20 % 11.99 %
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




eXTReMe Tracker