Method Details


Details for method 'MoKu Lab'

 

Method overview

name MoKu Lab
challenge pixel-level semantic labeling
details Alibaba, MoKu AI Lab, CV Group
publication Anonymous
project page / code
used Cityscapes data fine annotations
used external data ImageNet, Mapillary
runtime n/a
subsampling no
submission date January, 2020
previous submissions

 

Average results

Metric Value
IoU Classes 84.3131
iIoU Classes 65.1125
IoU Categories 92.6436
iIoU Categories 83.099

 

Class results

Class IoU iIoU
road 98.9093 -
sidewalk 88.7337 -
building 94.4419 -
wall 65.6609 -
fence 68.3631 -
pole 73.9481 -
traffic light 80.3206 -
traffic sign 83.7851 -
vegetation 94.3866 -
terrain 74.4692 -
sky 96.2413 -
person 88.8468 74.639
rider 76.2009 55.8934
car 96.7041 92.4232
truck 81.3474 50.5832
bus 93.1108 62.9904
train 91.5799 62.5554
motorcycle 74.7831 53.7011
bicycle 80.1152 68.114

 

Category results

Category IoU iIoU
flat 98.8537 -
nature 94.1382 -
object 79.2518 -
sky 96.2413 -
construction 94.7391 -
human 88.9547 75.4888
vehicle 96.3266 90.7092

 

Links

Download results as .csv file

Benchmark page