Method Details


Details for method 'Seg_UJS'

 

Method overview

name Seg_UJS
challenge pixel-level semantic labeling
details
publication Anonymous
project page / code
used Cityscapes data fine annotations
used external data ImageNet
runtime n/a
subsampling no
submission date September, 2020
previous submissions 1, 2, 3, 4

 

Average results

Metric Value
IoU Classes 84.2586
iIoU Classes 68.9877
IoU Categories 92.7305
iIoU Categories 84.3908

 

Class results

Class IoU iIoU
road 98.9676 -
sidewalk 89.2028 -
building 94.7512 -
wall 69.7285 -
fence 67.0694 -
pole 73.7012 -
traffic light 76.3994 -
traffic sign 82.1908 -
vegetation 94.4347 -
terrain 74.7589 -
sky 96.1653 -
person 89.6876 77.0989
rider 77.4153 60.6793
car 96.8933 91.9382
truck 81.2255 57.6256
bus 95.0882 70.5968
train 86.3809 62.0448
motorcycle 76.0297 61.2223
bicycle 80.8237 70.6959

 

Category results

Category IoU iIoU
flat 98.9154 -
nature 94.25 -
object 78.8026 -
sky 96.1653 -
construction 94.7993 -
human 89.6146 77.9266
vehicle 96.5662 90.8549

 

Links

Download results as .csv file

Benchmark page