Method Details
Details for method 'LDN-121'
Method overview
| name | LDN-121 |
| challenge | pixel-level semantic labeling |
| details | Ladder DenseNet-121 trained on train+val, fine labels only. Single-scale inference. |
| publication | Efficient Ladder-style DenseNets for Semantic Segmentation of Large Images Ivan Kreso, Josip Krapac, Sinisa Segvic |
| project page / code | |
| used Cityscapes data | fine annotations |
| used external data | ImageNet |
| runtime | 0.048 s Titan Xp |
| subsampling | no |
| submission date | November, 2018 |
| previous submissions |
Average results
| Metric | Value |
|---|---|
| IoU Classes | 79.3324 |
| iIoU Classes | 54.6514 |
| IoU Categories | 90.6872 |
| iIoU Categories | 78.3689 |
Class results
| Class | IoU | iIoU |
|---|---|---|
| road | 98.5756 | - |
| sidewalk | 85.7844 | - |
| building | 93.1336 | - |
| wall | 58.0861 | - |
| fence | 59.3712 | - |
| pole | 66.0433 | - |
| traffic light | 74.1095 | - |
| traffic sign | 77.6975 | - |
| vegetation | 93.4942 | - |
| terrain | 71.501 | - |
| sky | 95.6594 | - |
| person | 85.706 | 67.2539 |
| rider | 69.3867 | 45.1931 |
| car | 95.8925 | 89.9977 |
| truck | 71.2801 | 38.1393 |
| bus | 86.4959 | 46.6913 |
| train | 83.5224 | 47.9513 |
| motorcycle | 66.396 | 41.1216 |
| bicycle | 75.1806 | 60.8627 |
Category results
| Category | IoU | iIoU |
|---|---|---|
| flat | 98.6893 | - |
| nature | 93.1646 | - |
| object | 72.4881 | - |
| sky | 95.6594 | - |
| construction | 93.3384 | - |
| human | 86.0968 | 68.6302 |
| vehicle | 95.374 | 88.1076 |