Details for method 'GLNet_fine'
Method overview
| name |
GLNet_fine |
| challenge |
pixel-level semantic labeling |
| details |
The proposed network architecture, combined with spatial information and multi scale context information, and repair the boundaries and details of the segmented object through channel attention modules.(Use the train-fine and the val-fine data) |
| publication |
Anonymous
|
| project page / code |
|
| used Cityscapes data |
fine annotations |
| used external data |
ImageNet |
| runtime |
n/a |
| subsampling |
no |
| submission date |
June, 2019 |
| previous submissions |
|
Average results
| Metric |
Value |
| IoU Classes | 80.7776 |
| iIoU Classes | 58.2898 |
| IoU Categories | 91.3047 |
| iIoU Categories | 79.4568 |
Class results
| Class |
IoU |
iIoU |
| road | 98.6637 | - |
| sidewalk | 86.6852 | - |
| building | 93.448 | - |
| wall | 56.9209 | - |
| fence | 60.475 | - |
| pole | 68.2743 | - |
| traffic light | 75.4983 | - |
| traffic sign | 79.828 | - |
| vegetation | 93.6525 | - |
| terrain | 72.5765 | - |
| sky | 95.8544 | - |
| person | 87.0237 | 69.3537 |
| rider | 71.6354 | 49.988 |
| car | 96.0119 | 90.5069 |
| truck | 73.4582 | 39.7157 |
| bus | 90.5167 | 54.6211 |
| train | 85.744 | 52.7097 |
| motorcycle | 71.1131 | 47.3361 |
| bicycle | 77.3951 | 62.0871 |
Category results
| Category |
IoU |
iIoU |
| flat | 98.7309 | - |
| nature | 93.3612 | - |
| object | 74.3641 | - |
| sky | 95.8544 | - |
| construction | 93.832 | - |
| human | 87.3151 | 70.3822 |
| vehicle | 95.6753 | 88.5315 |
Links
Download results as .csv file
Benchmark page
