Method Details


Details for method 'DWGSeg-L75'

 

Method overview

name DWGSeg-L75
challenge pixel-level semantic labeling
details
publication Anonymous
project page / code
used Cityscapes data fine annotations
used external data
runtime 0.00755 s
1080
subsampling 1.33
submission date November, 2023
previous submissions

 

Average results

Metric Value
IoU Classes 76.6514
iIoU Classes 50.2722
IoU Categories 89.3807
iIoU Categories 75.0406

 

Class results

Class IoU iIoU
road 98.5188 -
sidewalk 85.4569 -
building 92.4087 -
wall 54.307 -
fence 55.5989 -
pole 61.5509 -
traffic light 70.5266 -
traffic sign 74.2721 -
vegetation 92.8837 -
terrain 70.7481 -
sky 95.1663 -
person 83.4861 62.5205
rider 66.6472 42.6512
car 95.3768 88.516
truck 71.6745 29.3399
bus 82.104 44.1014
train 72.0223 39.8562
motorcycle 61.1438 38.8591
bicycle 72.4835 56.3331

 

Category results

Category IoU iIoU
flat 98.5449 -
nature 92.5714 -
object 68.383 -
sky 95.1663 -
construction 92.6736 -
human 83.5993 63.8417
vehicle 94.7264 86.2394

 

Links

Download results as .csv file

Benchmark page