Method Details
Details for method 'F2MF-mid'
Method overview
| name | F2MF-mid |
| challenge | pixel-level semantic labeling |
| details | Our method forecasts semantic segmentation 9 timesteps into the future. |
| publication | Warp to the Future: Joint Forecasting of Features and Feature Motion Josip Saric, Marin Orsic, Tonci Antunovic, Sacha Vrazic, Sinisa Segvic The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020 http://openaccess.thecvf.com/content_CVPR_2020/html/Saric_Warp_to_the_Future_Joint_Forecasting_of_Features_and_Feature_CVPR_2020_paper.html |
| project page / code | https://jsaric.github.io/f2mf/ |
| used Cityscapes data | fine annotations, video |
| used external data | ImageNet |
| runtime | n/a |
| subsampling | no |
| submission date | November, 2019 |
| previous submissions |
Average results
| Metric | Value |
|---|---|
| IoU Classes | 59.1418 |
| iIoU Classes | 32.9474 |
| IoU Categories | 72.3534 |
| iIoU Categories | 47.9723 |
Class results
| Class | IoU | iIoU |
|---|---|---|
| road | 95.1334 | - |
| sidewalk | 69.1831 | - |
| building | 83.4636 | - |
| wall | 47.1769 | - |
| fence | 43.8126 | - |
| pole | 22.9257 | - |
| traffic light | 41.8166 | - |
| traffic sign | 41.3293 | - |
| vegetation | 84.1868 | - |
| terrain | 58.5158 | - |
| sky | 85.9859 | - |
| person | 46.6717 | 31.7294 |
| rider | 33.9405 | 17.0744 |
| car | 80.2939 | 65.4901 |
| truck | 53.7925 | 24.1308 |
| bus | 72.4972 | 34.3202 |
| train | 79.0464 | 41.8439 |
| motorcycle | 39.6563 | 20.6434 |
| bicycle | 44.265 | 28.3474 |
Category results
| Category | IoU | iIoU |
|---|---|---|
| flat | 95.2601 | - |
| nature | 83.6841 | - |
| object | 32.3701 | - |
| sky | 85.9859 | - |
| construction | 83.417 | - |
| human | 46.3418 | 31.9641 |
| vehicle | 79.4149 | 63.9805 |