Method

Superpixel-based Semantic Road Segmentation [SP-SS]
[Anonymous Submission]

Submitted on 12 Apr. 2017 15:52 by
[Anonymous Submission]

Running time:0.01 s
Environment:4 cores @ 3.0 Ghz (Matlab + C/C++)

Method Description:
The problem of fast and accurate road
segmentation problem is solved by proposing an
approach based on super-pixel segmentation and
Convolutional Neural Network. In this work an
irregular superpixels with regular grid
projection, for convolutional purpose,
embedding to higher dimensional feature space
is given as the input data model into the CNN
network.
Parameters:
learning rate=0.1*10^-4
weight decay=0.0001
Momentum=0.9
mini-batches size= 50
Latex Bibtex:

Evaluation in Bird's Eye View


Benchmark MaxF AP PRE REC FPR FNR
UM_ROAD 81.60 % 69.62 % 78.13 % 85.40 % 10.89 % 14.60 %
UMM_ROAD 85.07 % 79.86 % 85.97 % 84.20 % 15.11 % 15.80 %
UU_ROAD 78.47 % 65.18 % 74.20 % 83.25 % 9.43 % 16.75 %
URBAN_ROAD 82.36 % 72.31 % 80.48 % 84.33 % 11.27 % 15.67 %
This table as LaTeX

Behavior Evaluation


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