Method Details


Details for method 'SegModel'

 

Method overview

name SegModel
challenge pixel-level semantic labeling
details
publication Anonymous
project page / code
used Cityscapes data fine annotations
used external data ImageNet, MSCOCO
runtime 0.3 s
GTX TITAN X
subsampling no
submission date October, 2016
previous submissions 1

 

Average results

Metric Value
IoU Classes 77.6504
iIoU Classes 55.7493
IoU Categories 90.3627
iIoU Categories 77.5407

 

Class results

Class IoU iIoU
road 98.4914 -
sidewalk 85.4183 -
building 92.7933 -
wall 47.8581 -
fence 56.4756 -
pole 64.8097 -
traffic light 73.8459 -
traffic sign 77.943 -
vegetation 93.2803 -
terrain 71.4159 -
sky 95.5269 -
person 85.4662 66.0892
rider 68.9229 47.7858
car 95.7475 90.0391
truck 73.2998 40.4594
bus 81.3623 54.4557
train 69.1686 43.2961
motorcycle 68.6937 45.2709
bicycle 74.8375 58.5981

 

Category results

Category IoU iIoU
flat 98.6362 -
nature 93.0156 -
object 71.4794 -
sky 95.5269 -
construction 93.2468 -
human 85.5725 67.2096
vehicle 95.0619 87.8718

 

Links

Download results as .csv file

Benchmark page