Method Details


Details for method 'SR-AIC'

 

Method overview

name SR-AIC
challenge pixel-level semantic labeling
details
publication Anonymous
project page / code
used Cityscapes data fine annotations, coarse annotations
used external data ImageNet
runtime n/a
subsampling no
submission date November, 2017
previous submissions

 

Average results

Metric Value
IoU Classes 81.8882
iIoU Classes 60.6552
IoU Categories 91.3423
iIoU Categories 79.594

 

Class results

Class IoU iIoU
road 98.7155 -
sidewalk 87.1617 -
building 93.6677 -
wall 62.6356 -
fence 64.705 -
pole 68.9867 -
traffic light 76.4399 -
traffic sign 80.8141 -
vegetation 93.6802 -
terrain 73.284 -
sky 95.5454 -
person 86.8187 69.3119
rider 72.1753 51.2619
car 96.1975 90.6645
truck 77.9365 46.998
bus 90.6324 57.7449
train 87.8949 54.1838
motorcycle 71.2458 50.8676
bicycle 77.3383 64.2089

 

Category results

Category IoU iIoU
flat 98.7496 -
nature 93.4149 -
object 75.1001 -
sky 95.5454 -
construction 93.9541 -
human 86.8957 70.2193
vehicle 95.736 88.9686

 

Links

Download results as .csv file

Benchmark page