The KITTI Vision Benchmark Suite

Method

NIM-RTFNet [NIM-RTFNet]

Submitted on 4 Feb. 2020 14:16 by
Rui Fan (Tongji University)

Running time:		0.05 s
Environment:		GPU @ 2.5 Ghz (Python)

Method Description:

TBA

Parameters:

TBA

Latex Bibtex:

@InProceedings{wang2020applying,
author = {Wang, Hengli and Fan, Rui and
Sun, Yuxiang and Liu, Ming},
title = {Applying Surface Normal
Information in Drivable Area and Road Anomaly
Detection for Ground Mobile Robots},
booktitle = {2020 IEEE/RSJ International
Conference on Intelligent Robots and Systems
(IROS)},
year = {2020},
}

Evaluation in Bird's Eye View

Benchmark	MaxF	AP	PRE	REC	FPR	FNR
UM_ROAD	95.71 %	93.56 %	95.84 %	95.59 %	1.89 %	4.41 %
UMM_ROAD	96.79 %	95.61 %	97.03 %	96.54 %	3.25 %	3.46 %
UU_ROAD	95.11 %	92.94 %	95.91 %	94.32 %	1.31 %	5.68 %
URBAN_ROAD	96.02 %	94.01 %	96.43 %	95.62 %	1.95 %	4.38 %

This table as LaTeX

Behavior Evaluation

Benchmark	PRE-20	F1-20	HR-20	PRE-30	F1-30	HR-30	PRE-40	F1-40	HR-40

This table as LaTeX

Road/Lane Detection

The following plots show precision/recall curves for the bird's eye view evaluation.

This figure as: png eps pdf

Distance-dependent Behavior Evaluation

The following plots show the F1 score/Precision/Hitrate with respect to the longitudinal distance which has been used for evaluation.

Visualization of Results

The following images illustrate the performance of the method qualitatively on a couple of test images. We first show results in the perspective image, followed by evaluation in bird's eye view. Here, red denotes false negatives, blue areas correspond to false positives and green represents true positives.

This figure as: png