Method Details
Details for method 'ERINet-v2'
Method overview
| name | ERINet-v2 |
| challenge | pixel-level semantic labeling |
| details | |
| publication | Efficient Residual Inception Network MINJONG KIM, SUYOUNG CHI ongoing |
| project page / code | |
| used Cityscapes data | fine annotations |
| used external data | |
| runtime | 0.00526316 s 1080Ti, TensorRT v5.1.5 |
| subsampling | no |
| submission date | March, 2020 |
| previous submissions |
Average results
| Metric | Value |
|---|---|
| IoU Classes | 67.4063 |
| iIoU Classes | 39.2424 |
| IoU Categories | 85.7417 |
| iIoU Categories | 70.0373 |
Class results
| Class | IoU | iIoU |
|---|---|---|
| road | 97.7068 | - |
| sidewalk | 80.6235 | - |
| building | 89.6303 | - |
| wall | 41.7048 | - |
| fence | 45.0349 | - |
| pole | 52.1714 | - |
| traffic light | 55.9454 | - |
| traffic sign | 62.854 | - |
| vegetation | 91.482 | - |
| terrain | 69.1222 | - |
| sky | 93.8869 | - |
| person | 75.9157 | 55.3623 |
| rider | 53.7189 | 29.5132 |
| car | 92.2593 | 85.3261 |
| truck | 41.7945 | 17.9723 |
| bus | 66.3061 | 31.5529 |
| train | 64.2308 | 26.089 |
| motorcycle | 44.27 | 22.3264 |
| bicycle | 62.063 | 45.7972 |
Category results
| Category | IoU | iIoU |
|---|---|---|
| flat | 98.176 | - |
| nature | 91.1916 | - |
| object | 58.9205 | - |
| sky | 93.8869 | - |
| construction | 89.8489 | - |
| human | 76.576 | 56.786 |
| vehicle | 91.5919 | 83.2887 |