Method

MemRoadNet: Human-Like Memory Integration for Free Road Space Detection [MemRoadNet]
https://github.com/abdkhanstd/MRN

Submitted on 16 Aug. 2025 17:24 by
Abdullah Khan (Power labs)

Running time:0.01 s
Environment:GPU @ 2.5 Ghz (Python)

Method Description:
Human-Like Memory Integration for Free Road Space
Detection
Parameters:
None
Latex Bibtex:
Will update soon

Evaluation in Bird's Eye View


Benchmark MaxF AP PRE REC FPR FNR
UM_ROAD 96.55 % 93.58 % 96.46 % 96.64 % 1.61 % 3.36 %
UMM_ROAD 97.46 % 95.57 % 97.07 % 97.86 % 3.25 % 2.14 %
UU_ROAD 95.37 % 92.76 % 95.08 % 95.66 % 1.61 % 4.34 %
URBAN_ROAD 96.66 % 93.95 % 96.46 % 96.87 % 1.96 % 3.13 %
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

This figure as: png eps pdf

This figure as: png eps pdf

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

This figure as: png

This figure as: png

This figure as: png

This figure as: png

This figure as: png

This figure as: png

This figure as: png

This figure as: png

This figure as: png

This figure as: png

This figure as: png

This figure as: png

This figure as: png

This figure as: png

This figure as: png

This figure as: png

This figure as: png


eXTReMe Tracker