Method Details
Details for method 'SkipNet-MobileNet'
Method overview
| name | SkipNet-MobileNet |
| challenge | pixel-level semantic labeling |
| details | An efficient realtime semantic segmentation network with skip connections based on MobileNet. |
| publication | RTSeg: Real-time Semantic Segmentation Framework Mennatullah Siam, Mostafa Gamal, Moemen Abdel-Razek, Senthil Yogamani, Martin Jagersand Under Review by ICIP 2018 |
| project page / code | |
| used Cityscapes data | fine annotations, coarse annotations |
| used external data | ImageNet |
| runtime | n/a |
| subsampling | no |
| submission date | February, 2018 |
| previous submissions |
Average results
| Metric | Value |
|---|---|
| IoU Classes | 61.5187 |
| iIoU Classes | 35.1625 |
| IoU Categories | 82.0049 |
| iIoU Categories | 63.0336 |
Class results
| Class | IoU | iIoU |
|---|---|---|
| road | 95.8212 | - |
| sidewalk | 73.9302 | - |
| building | 86.1912 | - |
| wall | 36.6672 | - |
| fence | 39.3956 | - |
| pole | 44.5331 | - |
| traffic light | 47.244 | - |
| traffic sign | 54.339 | - |
| vegetation | 89.537 | - |
| terrain | 66.0146 | - |
| sky | 92.8904 | - |
| person | 69.2514 | 45.3982 |
| rider | 45.0519 | 26.0624 |
| car | 89.8774 | 80.1466 |
| truck | 35.6011 | 17.5923 |
| bus | 53.9098 | 27.572 |
| train | 45.615 | 23.4945 |
| motorcycle | 44.8316 | 22.1059 |
| bicycle | 58.1534 | 38.9283 |
Category results
| Category | IoU | iIoU |
|---|---|---|
| flat | 95.8675 | - |
| nature | 89.0597 | - |
| object | 51.4782 | - |
| sky | 92.8904 | - |
| construction | 85.9808 | - |
| human | 70.4409 | 47.6219 |
| vehicle | 88.3165 | 78.4453 |