Method Details
Details for method 'GPNet-ResNet101'
Method overview
name | GPNet-ResNet101 |
challenge | pixel-level semantic labeling |
details | |
publication | Anonymous |
project page / code | |
used Cityscapes data | fine annotations |
used external data | |
runtime | n/a |
subsampling | no |
submission date | February, 2020 |
previous submissions |
Average results
Metric | Value |
---|---|
IoU Classes | 82.4633 |
iIoU Classes | 62.3614 |
IoU Categories | 91.9263 |
iIoU Categories | 81.9876 |
Class results
Class | IoU | iIoU |
---|---|---|
road | 98.8367 | - |
sidewalk | 87.8891 | - |
building | 93.8945 | - |
wall | 61.8985 | - |
fence | 63.3048 | - |
pole | 70.4207 | - |
traffic light | 78.9208 | - |
traffic sign | 81.7752 | - |
vegetation | 94.0266 | - |
terrain | 72.4847 | - |
sky | 95.9278 | - |
person | 88.2967 | 73.0882 |
rider | 74.8748 | 54.175 |
car | 96.492 | 91.6648 |
truck | 80.4885 | 46.7671 |
bus | 91.1095 | 59.1435 |
train | 85.4575 | 54.5148 |
motorcycle | 72.0877 | 52.2051 |
bicycle | 78.6176 | 67.3327 |
Category results
Category | IoU | iIoU |
---|---|---|
flat | 98.7626 | - |
nature | 93.7003 | - |
object | 76.4405 | - |
sky | 95.9278 | - |
construction | 94.2254 | - |
human | 88.3677 | 74.0034 |
vehicle | 96.0595 | 89.9718 |