Method

Multi-branch Semantic Segmentation Network for Road Detection [MBRoadSeg]
[Anonymous Submission]

Submitted on 31 May. 2022 11:53 by
[Anonymous Submission]

Running time:.007 s
Environment:GPU @ 1.5 Ghz (Python)

Method Description:
The model separates the encoding process of road
context, local details and road boundary information
through a multi-branch structure shared by shallow
features. And each branch is carefully designed with
less computational overhead.
Parameters:
Epochs=100;Optimizer=Adam;
Latex Bibtex:

Evaluation in Bird's Eye View


Benchmark MaxF AP PRE REC FPR FNR
UM_ROAD 94.92 % 89.56 % 95.38 % 94.47 % 2.08 % 5.53 %
UMM_ROAD 96.88 % 92.93 % 96.98 % 96.78 % 3.31 % 3.22 %
UU_ROAD 94.19 % 87.69 % 94.00 % 94.39 % 1.96 % 5.61 %
URBAN_ROAD 95.64 % 90.30 % 95.78 % 95.51 % 2.32 % 4.49 %
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.


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