Method Details


Details for method 'MobileNetV2Plus'

 

Method overview

name MobileNetV2Plus
challenge pixel-level semantic labeling
details MobileNetV2Plus
publication Huijun Liu
project page / code
used Cityscapes data fine annotations
used external data ImageNet
runtime n/a
subsampling no
submission date March, 2018
previous submissions

 

Average results

Metric Value
IoU Classes 70.7195
iIoU Classes 46.772
IoU Categories 87.6404
iIoU Categories 72.8642

 

Class results

Class IoU iIoU
road 97.9747 -
sidewalk 81.9162 -
building 90.8729 -
wall 47.085 -
fence 48.5849 -
pole 57.129 -
traffic light 62.7919 -
traffic sign 68.2547 -
vegetation 92.2547 -
terrain 68.6211 -
sky 94.4821 -
person 80.2727 59.8022
rider 60.434 40.8573
car 93.833 85.7053
truck 59.099 30.2503
bus 65.3518 37.0664
train 55.7318 31.2799
motorcycle 51.963 34.3961
bicycle 67.0174 54.8182

 

Category results

Category IoU iIoU
flat 98.3693 -
nature 91.882 -
object 63.9853 -
sky 94.4821 -
construction 91.1946 -
human 80.4392 61.8257
vehicle 93.1299 83.9026

 

Links

Download results as .csv file

Benchmark page