Method

Sparse Query Dense: Enhancing 3D Object Detection with Pseudo Points [SQD]
https://github.com/yujmo/SQDNet

Submitted on 24 Mar. 2024 13:11 by
mo yujian (tongji university)

Running time:0.06 s
Environment:1 core @ 2.5 Ghz (C/C++)

Method Description:
Use synthetic data to supplement real data.
Parameters:
10,3,0.9,5,76
Latex Bibtex:
@inproceedings{Yujian_Mo_acmmm2024,
author = {Yujian Mo and Yan Wu and Junqiao Zhao and Zhenjie Hou and Weiquan Huang and Yinghao Hu and Jijun Wang and Jun Yan},
title = {Sparse Query Dense: Enhancing 3D Object Detection with Pseudo Points},
booktitle = {ACM MM Oral},
year = {2024},
address = {Melbourne, VIC, Australia},
publisher = {ACM},
doi = {10.1145/3664647.3681420}
}

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) 98.21 % 94.92 % 92.37 %
Car (Orientation) 98.20 % 94.85 % 92.26 %
Car (3D Detection) 91.58 % 81.82 % 79.07 %
Car (Bird's Eye View) 95.44 % 90.63 % 88.04 %
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