Method Details


Details for method 'TUE-5LSM0-g23'

 

Method overview

name TUE-5LSM0-g23
challenge pixel-level semantic labeling
details Deeplabv3+decoder
publication Anonymous
project page / code
used Cityscapes data fine annotations, coarse annotations
used external data ImageNet
runtime n/a
subsampling no
submission date April, 2020
previous submissions

 

Average results

Metric Value
IoU Classes 65.8303
iIoU Classes 42.4327
IoU Categories 81.829
iIoU Categories 67.8274

 

Class results

Class IoU iIoU
road 95.4744 -
sidewalk 77.7202 -
building 85.8736 -
wall 45.7467 -
fence 48.2543 -
pole 43.9226 -
traffic light 53.1712 -
traffic sign 57.5177 -
vegetation 89.2331 -
terrain 65.2954 -
sky 91.0526 -
person 70.7305 53.8417
rider 54.2053 32.5941
car 91.2806 82.946
truck 48.1044 23.4382
bus 67.5306 37.8429
train 53.514 32.9334
motorcycle 51.7825 28.7921
bicycle 60.3657 47.0736

 

Category results

Category IoU iIoU
flat 95.702 -
nature 88.8365 -
object 51.5118 -
sky 91.0526 -
construction 85.6921 -
human 70.6436 54.6645
vehicle 89.3643 80.9903

 

Links

Download results as .csv file

Benchmark page