Method Details
Details for method 'ResNeSt200'
Method overview
| name | ResNeSt200 |
| challenge | pixel-level semantic labeling |
| details | DeepLabV3+ network with ResNeSt200 backbone. |
| publication | ResNeSt: Split-Attention Networks Hang Zhang, Chongruo Wu, Zhongyue Zhang, Yi Zhu, Zhi Zhang, Haibin Lin, Yue Sun, Tong He, Jonas Mueller, R. Manmatha, Mu Li, and Alexander Smola https://arxiv.org/abs/2004.08955 |
| project page / code | |
| used Cityscapes data | fine annotations |
| used external data | ImageNet, Mapillary |
| runtime | n/a |
| subsampling | no |
| submission date | May, 2020 |
| previous submissions |
Average results
| Metric | Value |
|---|---|
| IoU Classes | 83.2611 |
| iIoU Classes | 63.0098 |
| IoU Categories | 92.3056 |
| iIoU Categories | 81.2647 |
Class results
| Class | IoU | iIoU |
|---|---|---|
| road | 98.8882 | - |
| sidewalk | 88.4154 | - |
| building | 94.3537 | - |
| wall | 65.9747 | - |
| fence | 66.0136 | - |
| pole | 72.543 | - |
| traffic light | 78.5787 | - |
| traffic sign | 82.4758 | - |
| vegetation | 94.1842 | - |
| terrain | 72.864 | - |
| sky | 96.2613 | - |
| person | 88.4383 | 71.9145 |
| rider | 74.8105 | 53.7176 |
| car | 96.5826 | 91.7784 |
| truck | 77.0118 | 47.1849 |
| bus | 92.3494 | 63.4171 |
| train | 89.9938 | 58.8968 |
| motorcycle | 73.1583 | 52.1416 |
| bicycle | 79.0634 | 65.0273 |
Category results
| Category | IoU | iIoU |
|---|---|---|
| flat | 98.8009 | - |
| nature | 93.8627 | - |
| object | 77.8493 | - |
| sky | 96.2613 | - |
| construction | 94.5723 | - |
| human | 88.5479 | 72.6212 |
| vehicle | 96.2446 | 89.9083 |