Method Details


Details for method 'motovis'

 

Method overview

name motovis
challenge pixel-level semantic labeling
details
publication motovis.com
project page / code
used Cityscapes data fine annotations, coarse annotations
used external data ImageNet
runtime n/a
subsampling no
submission date May, 2017
previous submissions 1

 

Average results

Metric Value
IoU Classes 81.3256
iIoU Classes 57.6571
IoU Categories 91.4564
iIoU Categories 80.651

 

Class results

Class IoU iIoU
road 98.6622 -
sidewalk 86.5649 -
building 93.467 -
wall 55.4533 -
fence 62.6513 -
pole 69.3811 -
traffic light 76.2975 -
traffic sign 80.4474 -
vegetation 93.8468 -
terrain 72.6419 -
sky 95.7524 -
person 87.1008 71.3237
rider 72.3662 48.089
car 96.1804 91.0868
truck 77.8843 40.8614
bus 91.2559 50.7558
train 88.6331 50.5274
motorcycle 69.4877 46.5587
bicycle 77.1121 62.054

 

Category results

Category IoU iIoU
flat 98.7144 -
nature 93.4125 -
object 75.3561 -
sky 95.7524 -
construction 93.8357 -
human 87.3069 72.2899
vehicle 95.8168 89.0121

 

Links

Download results as .csv file

Benchmark page