KITTI-360

Method

DEVIANT: Depth Equivariant Network for Moncoular 3D Object Detection [DEVIANT]
https://github.com/abhi1kumar/DEVIANT

Submitted on 17 Apr. 2023 23:26 by
Andrew Hou (Michigan State University)

Running time:0.04 s
Environment:NVIDIA A100

Method Description:
DEVIANT (ECCV 2022) trained on KITTI-360 Train+Val
Parameters:
sesn_scales: [0.83, 1.0 ]
2 classes: Cars and Buildings
Latex Bibtex:
@inproceedings{kumar2022deviant,
title={{DEVIANT: Depth Equivariant Network for Monocular 3D Object Detection}},
author={Kumar, Abhinav and Brazil, Garrick and Corona, Enrique and Parchami, Armin and Liu, Xiaoming},
booktitle={ECCV},
year={2022}
}

Test Set Average

AP25% AP50%
26.96 0.88
This table as LaTeX




eXTReMe Tracker