Method Details


Details for method 'AntGroup-AI-VisionAlgo'

 

Method overview

name AntGroup-AI-VisionAlgo
challenge pixel-level semantic labeling
details AntGroup AI vision algo
publication Anonymous
project page / code
used Cityscapes data fine annotations, coarse annotations
used external data ImageNet, Mapillary
runtime n/a
subsampling no
submission date February, 2021
previous submissions 1, 2, 3

 

Average results

Metric Value
IoU Classes 86.0768
iIoU Classes 71.3153
IoU Categories 93.169
iIoU Categories 84.6696

 

Class results

Class IoU iIoU
road 98.8564 -
sidewalk 88.7003 -
building 95.0358 -
wall 74.1813 -
fence 71.1869 -
pole 76.2949 -
traffic light 82.0552 -
traffic sign 84.851 -
vegetation 94.57 -
terrain 75.1504 -
sky 96.2946 -
person 90.0094 77.3702
rider 79.8129 63.3336
car 96.8228 92.1176
truck 83.728 62.6484
bus 94.9137 71.2365
train 93.4274 68.6398
motorcycle 77.7488 63.3928
bicycle 81.8198 71.7832

 

Category results

Category IoU iIoU
flat 98.8659 -
nature 94.3325 -
object 81.0416 -
sky 96.2946 -
construction 95.0573 -
human 90.0074 78.1545
vehicle 96.584 91.1846

 

Links

Download results as .csv file

Benchmark page