Details for method 'DDAR'
Method overview
| name |
DDAR |
| challenge |
pixel-level semantic labeling |
| details |
DiDi Labs, AR Group |
| publication |
Anonymous
|
| project page / code |
|
| used Cityscapes data |
fine annotations, coarse annotations |
| used external data |
ImageNet |
| runtime |
n/a |
| subsampling |
no |
| submission date |
November, 2018 |
| previous submissions |
|
Average results
| Metric |
Value |
| IoU Classes | 82.231 |
| iIoU Classes | 62.6874 |
| IoU Categories | 91.8735 |
| iIoU Categories | 81.5498 |
Class results
| Class |
IoU |
iIoU |
| road | 98.7461 | - |
| sidewalk | 87.2726 | - |
| building | 93.6221 | - |
| wall | 57.1833 | - |
| fence | 62.8903 | - |
| pole | 70.9172 | - |
| traffic light | 77.7919 | - |
| traffic sign | 82.0498 | - |
| vegetation | 93.8121 | - |
| terrain | 71.4386 | - |
| sky | 95.7589 | - |
| person | 88.0935 | 72.9565 |
| rider | 74.4998 | 54.4711 |
| car | 96.3227 | 90.8436 |
| truck | 80.13 | 47.6799 |
| bus | 91.3239 | 58.7453 |
| train | 87.5514 | 57.543 |
| motorcycle | 73.9569 | 52.5794 |
| bicycle | 79.0285 | 66.6802 |
Category results
| Category |
IoU |
iIoU |
| flat | 98.7403 | - |
| nature | 93.5307 | - |
| object | 76.7469 | - |
| sky | 95.7589 | - |
| construction | 94.09 | - |
| human | 88.3365 | 74.0095 |
| vehicle | 95.9112 | 89.0901 |
Links
Download results as .csv file
Benchmark page
