Method Details


Details for method 'MCANet'

 

Method overview

name MCANet
challenge pixel-level semantic labeling
details
publication Anonymous
project page / code https://github.com/tanmaysingha/MCANet
used Cityscapes data fine annotations, coarse annotations
used external data
runtime n/a
subsampling no
submission date September, 2021
previous submissions

 

Average results

Metric Value
IoU Classes 73.4351
iIoU Classes 45.7823
IoU Categories 88.9269
iIoU Categories 72.8328

 

Class results

Class IoU iIoU
road 98.3279 -
sidewalk 84.1989 -
building 91.9989 -
wall 49.4585 -
fence 51.5547 -
pole 62.0857 -
traffic light 67.7875 -
traffic sign 73.1866 -
vegetation 92.5588 -
terrain 70.252 -
sky 95.34 -
person 81.4057 58.5425
rider 59.2058 36.2088
car 94.5947 87.7211
truck 57.8211 27.2026
bus 75.9103 41.5234
train 64.169 29.7082
motorcycle 56.323 32.6482
bicycle 69.0876 52.7038

 

Category results

Category IoU iIoU
flat 98.4695 -
nature 92.2257 -
object 68.3856 -
sky 95.34 -
construction 91.9684 -
human 82.2505 60.0857
vehicle 93.8485 85.5799

 

Links

Download results as .csv file

Benchmark page