Method

[la]SE-SSD: Self-Ensembling Single-Stage Object Detector From Point Cloud [SE-SSD]
https://github.com/Vegeta2020/SE-SSD

Submitted on 14 Nov. 2020 13:32 by
Weiliang Tang (CUHK)

Running time:0.03 s
Environment:1 core @ 2.5 Ghz (Python + C/C++)

Method Description:
TBA
Parameters:
TBA
Latex Bibtex:
@inproceedings{zheng2020ciassd,
title={SE-SSD: Self-Ensembling Single-Stage Object
Detector From Point Cloud},
author={Zheng, Wu and Tang, Weiliang and Jiang, Li
and Fu, Chi-Wing},
booktitle={CVPR},
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) 96.69 % 95.60 % 90.53 %
Car (Orientation) 96.55 % 95.17 % 90.00 %
Car (3D Detection) 91.49 % 82.54 % 77.15 %
Car (Bird's Eye View) 95.68 % 91.84 % 86.72 %
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