Details for method 'DeepMotion'
Method overview
| name |
DeepMotion |
| challenge |
pixel-level semantic labeling |
| details |
We propose a novel method based on convnets to extract multi-scale features in a large range particularly for solving street scene segmentation. |
| publication |
Anonymous
|
| project page / code |
|
| used Cityscapes data |
fine annotations |
| used external data |
|
| runtime |
n/a |
| subsampling |
no |
| submission date |
November, 2017 |
| previous submissions |
|
Average results
| Metric |
Value |
| IoU Classes | 81.3525 |
| iIoU Classes | 58.5866 |
| IoU Categories | 90.6948 |
| iIoU Categories | 78.1168 |
Class results
| Class |
IoU |
iIoU |
| road | 98.7071 | - |
| sidewalk | 87.0493 | - |
| building | 93.4582 | - |
| wall | 61.6127 | - |
| fence | 62.5517 | - |
| pole | 65.3846 | - |
| traffic light | 74.5588 | - |
| traffic sign | 78.6379 | - |
| vegetation | 93.6098 | - |
| terrain | 72.549 | - |
| sky | 95.4174 | - |
| person | 86.1694 | 67.4709 |
| rider | 72.2531 | 49.1821 |
| car | 96.1047 | 89.8858 |
| truck | 82.3225 | 44.1618 |
| bus | 92.818 | 55.2125 |
| train | 85.6624 | 55.6006 |
| motorcycle | 70.217 | 46.2726 |
| bicycle | 76.613 | 60.9067 |
Category results
| Category |
IoU |
iIoU |
| flat | 98.6711 | - |
| nature | 93.3026 | - |
| object | 72.1087 | - |
| sky | 95.4174 | - |
| construction | 93.6184 | - |
| human | 86.215 | 68.4803 |
| vehicle | 95.5301 | 87.7533 |
Links
Download results as .csv file
Benchmark page
