Anonymous Results


 

Method overview

challenge pixel-level semantic labeling
used Cityscapes data fine annotations
used external data
runtime n/a
subsampling no

 

Average results

Metric Value
IoU Classes 70.0688
iIoU Classes 42.8617
IoU Categories 87.8098
iIoU Categories 71.5814

 

Class results

Class IoU iIoU
road 96.8381 -
sidewalk 78.5015 -
building 90.9874 -
wall 45.3994 -
fence 50.1549 -
pole 59.1089 -
traffic light 65.2153 -
traffic sign 70.792 -
vegetation 92.5142 -
terrain 68.1722 -
sky 94.668 -
person 80.5683 58.6862
rider 58.5013 34.8481
car 92.7457 85.6664
truck 52.7479 25.7268
bus 67.2585 37.4329
train 50.9541 28.6264
motorcycle 50.4596 27.361
bicycle 65.719 44.5459

 

Category results

Category IoU iIoU
flat 97.9286 -
nature 92.06 -
object 65.6888 -
sky 94.668 -
construction 91.6702 -
human 80.8651 60.0942
vehicle 91.7878 83.0685

 

Links

Download results as .csv file

Benchmark page