Method

PV-DT3D [PV-DT3D]


Submitted on 27 Sep. 2022 09:28 by
帆航 杨 (天津理工大学)

Running time:1.4 s
Environment:1 core @ 2.5 Ghz (C/C++)

Method Description:
We proposes a new 3D object detection framework
called Point-Voxel dual transformer(PV-DT3D)
Parameters:
alpha=1.1
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) 96.34 % 95.36 % 92.70 %
Car (Orientation) 96.33 % 95.23 % 92.53 %
Car (3D Detection) 90.07 % 82.09 % 77.51 %
Car (Bird's Eye View) 92.65 % 89.10 % 86.43 %
This table as LaTeX

2D object detection results.
This figure as: png eps txt gnuplot



Orientation estimation results.
This figure as: png eps txt gnuplot



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



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




eXTReMe Tracker