Details for method 'NV-ADLR'
Method overview
| name |
NV-ADLR |
| challenge |
pixel-level semantic labeling |
| details |
NVIDIA Applied Deep Learning Research |
| publication |
Anonymous
|
| project page / code |
|
| used Cityscapes data |
fine annotations, coarse annotations |
| used external data |
ImageNet |
| runtime |
n/a |
| subsampling |
no |
| submission date |
April, 2018 |
| previous submissions |
1 |
Average results
| Metric |
Value |
| IoU Classes | 82.1681 |
| iIoU Classes | 63.1132 |
| IoU Categories | 91.8468 |
| iIoU Categories | 82.1938 |
Class results
| Class |
IoU |
iIoU |
| road | 98.6718 | - |
| sidewalk | 87.1044 | - |
| building | 93.8265 | - |
| wall | 60.8939 | - |
| fence | 63.9494 | - |
| pole | 70.9054 | - |
| traffic light | 78.0266 | - |
| traffic sign | 81.5275 | - |
| vegetation | 94.0861 | - |
| terrain | 74.3065 | - |
| sky | 96.0196 | - |
| person | 87.3782 | 73.3789 |
| rider | 72.6954 | 54.0825 |
| car | 96.2282 | 91.8232 |
| truck | 74.4984 | 46.4249 |
| bus | 91.4882 | 59.9079 |
| train | 89.5661 | 60.641 |
| motorcycle | 71.656 | 52.3476 |
| bicycle | 78.366 | 66.2999 |
Category results
| Category |
IoU |
iIoU |
| flat | 98.7781 | - |
| nature | 93.7548 | - |
| object | 76.7147 | - |
| sky | 96.0196 | - |
| construction | 94.1316 | - |
| human | 87.597 | 74.2416 |
| vehicle | 95.9321 | 90.146 |
Links
Download results as .csv file
Benchmark page
