Method Details


Details for method 'iFLYTEK-CV'

 

Method overview

name iFLYTEK-CV
challenge pixel-level semantic labeling
details iFLYTEK Research, CV Group
publication Anonymous
project page / code
used Cityscapes data fine annotations, coarse annotations
used external data ImageNet; Mapillary
runtime n/a
subsampling no
submission date October, 2018
previous submissions

 

Average results

Metric Value
IoU Classes 83.6099
iIoU Classes 64.7379
IoU Categories 92.0662
iIoU Categories 82.3274

 

Class results

Class IoU iIoU
road 98.6989 -
sidewalk 87.0938 -
building 94.1161 -
wall 64.3918 -
fence 65.3587 -
pole 71.2097 -
traffic light 77.8737 -
traffic sign 82.1547 -
vegetation 94.0232 -
terrain 73.4984 -
sky 95.9724 -
person 88.3168 73.9771
rider 75.7099 53.9754
car 96.5053 91.3714
truck 83.2729 52.3582
bus 94.6735 63.0407
train 92.3539 63.0679
motorcycle 74.2903 53.5015
bicycle 79.0733 66.6108

 

Category results

Category IoU iIoU
flat 98.7744 -
nature 93.703 -
object 76.9897 -
sky 95.9724 -
construction 94.2468 -
human 88.5208 74.7596
vehicle 96.256 89.8952

 

Links

Download results as .csv file

Benchmark page