Method Details

Details for method 'Instance-level Segmentation of Vehicles by Deep Contours'


Method overview

name Instance-level Segmentation of Vehicles by Deep Contours
challenge instance-level semantic labeling
details Our method uses the fully convolutional network (FCN) for semantic labeling and for estimating the boundary of each vehicle. Even though a contour is in general a one pixel wide structure which cannot be directly learned by a CNN, our network addresses this by providing areas around the contours. Based on these areas, we separate the individual vehicle instances.
publication Instance-level Segmentation of Vehicles by Deep Contours
Jan van den Brand, Matthias Ochs and Rudolf Mester
Asian Conference on Computer Vision - Workshop on Computer Vision Technologies for Smart Vehicle
project page / code
used Cityscapes data fine annotations
used external data ImageNet
runtime 0.2 s
GTX 1070
subsampling 2
submission date August, 2016
previous submissions


Average results

Metric Value
AP 2.26995
AP50% 3.6534
AP100m 3.87795
AP50m 4.87125


Class results

Class AP AP50% AP100m AP50m
person 0 0 0 0
rider 0 0 0 0
car 18.1596 29.2272 31.0236 38.97
truck 0 0 0 0
bus 0 0 0 0
train 0 0 0 0
motorcycle 0 0 0 0
bicycle 0 0 0 0



Download results as .csv file

Benchmark page