Method Details


Details for method 'VENUS_ROB'

 

Method overview

name VENUS_ROB
challenge pixel-level semantic labeling
details Robust Vision Challenge 2018
publication Anonymous
project page / code
used Cityscapes data fine annotations
used external data ImageNet, ADE20K, KITTI, WildDash, ScanNet
runtime 0.5 s
Nivida Titan X
subsampling no
submission date May, 2018
previous submissions 1

 

Average results

Metric Value
IoU Classes 65.2299
iIoU Classes 35.9235
IoU Categories 83.6835
iIoU Categories 64.5332

 

Class results

Class IoU iIoU
road 95.936 -
sidewalk 75.8226 -
building 88.9897 -
wall 45.8537 -
fence 42.7845 -
pole 41.0767 -
traffic light 53.5234 -
traffic sign 61.5113 -
vegetation 90.5439 -
terrain 66.5806 -
sky 93.4186 -
person 74.7125 47.9347
rider 47.7438 21.8352
car 92.4173 82.3782
truck 46.8966 19.8246
bus 58.567 30.1789
train 52.9209 23.8458
motorcycle 47.7464 20.5619
bicycle 62.3235 40.8289

 

Category results

Category IoU iIoU
flat 95.8224 -
nature 90.1889 -
object 50.3988 -
sky 93.4186 -
construction 88.8541 -
human 75.9335 49.4184
vehicle 91.1681 79.6481

 

Links

Download results as .csv file

Benchmark page