The KITTI Vision Benchmark Suite

Method

Color intuitive feature guided depth-height fusion and volume rendering for monocular 3D detection [MonoFRD]

Submitted on 27 Mar. 2024 02:46 by
zhipeng gong (csu)

Running time:		0.01 s
Environment:		1 core @ 2.5 Ghz (C/C++)

Method Description:

A monocular 3d object detection method.

Parameters:

TBD

Latex Bibtex:

@article{gong2024color,
author={Gong, Zhipeng and Zhao, Yuqian and Zhang,
Fan and Gui, Gui and Chen, Baifan and Yu, Lingli and
Wang, Hui and Yang, Chunhua and Gui, Weihua},
journal={IEEE Transactions on Intelligent
Vehicles(Major Revison)},
title={Color intuitive feature guided depth-height
fusion and volume rendering for monocular 3D object
detection},
year={2024},
publisher={IEEE},
doi={},
volume={},
number={},
pages={}
}

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)	95.25 %	87.31 %	77.66 %
Car (Orientation)	95.01 %	86.58 %	76.82 %
Car (3D Detection)	21.11 %	16.24 %	14.97 %
Car (Bird's Eye View)	29.65 %	22.77 %	20.41 %
Pedestrian (Detection)	54.06 %	41.20 %	37.53 %
Pedestrian (Orientation)	33.47 %	24.92 %	22.38 %
Pedestrian (3D Detection)	13.86 %	8.88 %	7.53 %
Pedestrian (Bird's Eye View)	15.68 %	10.38 %	8.79 %
Cyclist (Detection)	55.86 %	38.98 %	34.32 %
Cyclist (Orientation)	39.05 %	26.87 %	24.09 %
Cyclist (3D Detection)	6.38 %	3.33 %	3.12 %
Cyclist (Bird's Eye View)	8.44 %	4.55 %	4.14 %