Method Details


Details for method 'ARSAIT'

 

Method overview

name ARSAIT
challenge pixel-level semantic labeling
details anonymous
publication Anonymous
project page / code
used Cityscapes data fine annotations
used external data ImageNet
runtime 1 s
p40
subsampling no
submission date January, 2018
previous submissions 1

 

Average results

Metric Value
IoU Classes 73.5546
iIoU Classes 48.2466
IoU Categories 88.9706
iIoU Categories 74.7572

 

Class results

Class IoU iIoU
road 98.152 -
sidewalk 82.6893 -
building 91.8922 -
wall 48.5255 -
fence 51.3016 -
pole 60.6739 -
traffic light 67.7132 -
traffic sign 73.3775 -
vegetation 92.7822 -
terrain 69.9487 -
sky 95.154 -
person 82.624 61.7182
rider 61.9039 38.3593
car 94.8918 88.1758
truck 58.7856 27.6076
bus 66.7275 42.819
train 66.7064 33.6475
motorcycle 62.3185 36.9726
bicycle 71.3699 56.6723

 

Category results

Category IoU iIoU
flat 98.4667 -
nature 92.3606 -
object 67.6201 -
sky 95.154 -
construction 92.138 -
human 83.1622 63.1303
vehicle 93.8926 86.3842

 

Links

Download results as .csv file

Benchmark page