Method Details
Details for method 'SegModel'
Method overview
| name | SegModel |
| challenge | pixel-level semantic labeling |
| details | Both train set (2975) and val set (500) are used to train model for this submission. |
| publication | Anonymous |
| project page / code | |
| used Cityscapes data | fine annotations |
| used external data | ImageNet, MSCOCO |
| runtime | 0.8 s GTX TITAN X |
| subsampling | no |
| submission date | November, 2016 |
| previous submissions | 1, 2 |
Average results
| Metric | Value |
|---|---|
| IoU Classes | 78.4736 |
| iIoU Classes | 56.0544 |
| IoU Categories | 89.8238 |
| iIoU Categories | 75.9337 |
Class results
| Class | IoU | iIoU |
|---|---|---|
| road | 98.5764 | - |
| sidewalk | 86.3538 | - |
| building | 92.7854 | - |
| wall | 52.3698 | - |
| fence | 59.6527 | - |
| pole | 59.6398 | - |
| traffic light | 72.4771 | - |
| traffic sign | 78.2628 | - |
| vegetation | 93.2824 | - |
| terrain | 72.7833 | - |
| sky | 95.508 | - |
| person | 85.3506 | 63.3372 |
| rider | 70.05 | 47.0218 |
| car | 95.653 | 89.8406 |
| truck | 75.3836 | 41.4424 |
| bus | 84.1378 | 56.22 |
| train | 75.0596 | 48.4347 |
| motorcycle | 68.7214 | 45.4453 |
| bicycle | 74.9515 | 56.6929 |
Category results
| Category | IoU | iIoU |
|---|---|---|
| flat | 98.6285 | - |
| nature | 93.0199 | - |
| object | 68.1022 | - |
| sky | 95.508 | - |
| construction | 93.2687 | - |
| human | 85.2397 | 64.225 |
| vehicle | 94.9996 | 87.6424 |