Anonymous Results


 

Method overview

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

 

Average results

Metric Value
IoU Classes 84.7203
iIoU Classes 68.4453
IoU Categories 92.9372
iIoU Categories 84.2829

 

Class results

Class IoU iIoU
road 98.9383 -
sidewalk 89.1336 -
building 94.7428 -
wall 70.8278 -
fence 67.3797 -
pole 75.4731 -
traffic light 81.7913 -
traffic sign 84.8139 -
vegetation 94.3532 -
terrain 74.4036 -
sky 96.1855 -
person 89.5055 76.3341
rider 78.7167 61.9068
car 96.8671 92.3953
truck 79.8323 56.2092
bus 92.8322 67.962
train 86.9675 61.9653
motorcycle 76.1755 60.1186
bicycle 80.7452 70.6712

 

Category results

Category IoU iIoU
flat 98.8691 -
nature 94.0998 -
object 80.4948 -
sky 96.1855 -
construction 94.7773 -
human 89.6142 77.3421
vehicle 96.5199 91.2236

 

Links

Download results as .csv file

Benchmark page