Details for method 'LeapAI'
Method overview
| name |
LeapAI |
| challenge |
pixel-level semantic labeling |
| details |
Using advanced AI techniques. |
| publication |
Anonymous
|
| project page / code |
|
| used Cityscapes data |
fine annotations, coarse annotations |
| used external data |
ImageNet |
| runtime |
n/a |
| subsampling |
no |
| submission date |
April, 2022 |
| previous submissions |
|
Average results
| Metric |
Value |
| IoU Classes | 86.3918 |
| iIoU Classes | 70.9035 |
| IoU Categories | 93.2127 |
| iIoU Categories | 84.1736 |
Class results
| Class |
IoU |
iIoU |
| road | 99.0223 | - |
| sidewalk | 89.8536 | - |
| building | 95.0843 | - |
| wall | 73.7509 | - |
| fence | 71.5659 | - |
| pole | 76.0238 | - |
| traffic light | 82.4007 | - |
| traffic sign | 85.9037 | - |
| vegetation | 94.5778 | - |
| terrain | 76.0138 | - |
| sky | 96.3882 | - |
| person | 90.3393 | 76.6956 |
| rider | 80.3032 | 63.5393 |
| car | 97.0311 | 92.5812 |
| truck | 84.1747 | 62.9269 |
| bus | 95.0924 | 72.5944 |
| train | 92.1252 | 63.3906 |
| motorcycle | 79.2888 | 63.7716 |
| bicycle | 82.5046 | 71.7282 |
Category results
| Category |
IoU |
iIoU |
| flat | 98.9511 | - |
| nature | 94.3763 | - |
| object | 81.1619 | - |
| sky | 96.3882 | - |
| construction | 94.8847 | - |
| human | 90.0441 | 77.0738 |
| vehicle | 96.6828 | 91.2733 |
Links
Download results as .csv file
Benchmark page
