Method Details
Details for method 'SN_RN152pyrx8_RVC'
Method overview
| name | SN_RN152pyrx8_RVC |
| challenge | pixel-level semantic labeling |
| details | |
| publication | In Defense of Pre-trained ImageNet Architectures for Real-time Semantic Segmentation of Road-driving Images Marin Oršić, Ivan Krešo, Petra Bevandić, Siniša Šegvić CVPR 2019 https://openaccess.thecvf.com/content_CVPR_2019/papers/Orsic_In_Defense_of_Pre-Trained_ImageNet_Architectures_for_Real-Time_Semantic_Segmentation_CVPR_2019_paper.pdf |
| project page / code | https://github.com/orsic/swiftnet |
| used Cityscapes data | fine annotations |
| used external data | ImageNet, ADE20k, KITTI, MVD, ScanNet, VIPER, WildDash2 |
| runtime | 1 s Tesla V100 |
| subsampling | no |
| submission date | August, 2020 |
| previous submissions |
Average results
| Metric | Value |
|---|---|
| IoU Classes | 74.6506 |
| iIoU Classes | 50.5134 |
| IoU Categories | 89.4161 |
| iIoU Categories | 75.8903 |
Class results
| Class | IoU | iIoU |
|---|---|---|
| road | 98.4013 | - |
| sidewalk | 84.8028 | - |
| building | 92.5502 | - |
| wall | 55.999 | - |
| fence | 53.5996 | - |
| pole | 61.0845 | - |
| traffic light | 70.3039 | - |
| traffic sign | 74.1919 | - |
| vegetation | 93.0698 | - |
| terrain | 71.1436 | - |
| sky | 95.5119 | - |
| person | 82.8261 | 63.6821 |
| rider | 63.1495 | 36.592 |
| car | 95.3229 | 88.7686 |
| truck | 65.976 | 42.3318 |
| bus | 72.0111 | 47.8617 |
| train | 53.705 | 32.3249 |
| motorcycle | 63.0699 | 38.206 |
| bicycle | 71.6424 | 54.3402 |
Category results
| Category | IoU | iIoU |
|---|---|---|
| flat | 98.5472 | - |
| nature | 92.7235 | - |
| object | 68.0888 | - |
| sky | 95.5119 | - |
| construction | 92.7519 | - |
| human | 83.6689 | 65.1723 |
| vehicle | 94.6207 | 86.6083 |