Method Details
Details for method 'Segnet extended'
Method overview
| name | Segnet extended |
| challenge | pixel-level semantic labeling |
| details | Trained on a pre-release version of the dataset |
| publication | SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation V. Badrinarayanan, A. Kendall, and R. Cipolla arXiv preprint 2015 http://arxiv.org/pdf/1511.00561v2 |
| project page / code | https://github.com/alexgkendall/caffe-segnet |
| used Cityscapes data | fine annotations |
| used external data | ImageNet |
| runtime | 0.06 s |
| subsampling | 4 |
| submission date | April, 2016 |
| previous submissions |
Average results
| Metric | Value |
|---|---|
| IoU Classes | 56.0937 |
| iIoU Classes | 34.2066 |
| IoU Categories | 79.8263 |
| iIoU Categories | 66.4144 |
Class results
| Class | IoU | iIoU |
|---|---|---|
| road | 95.6463 | - |
| sidewalk | 70.1007 | - |
| building | 82.8084 | - |
| wall | 29.8726 | - |
| fence | 31.8846 | - |
| pole | 38.062 | - |
| traffic light | 43.0533 | - |
| traffic sign | 44.5793 | - |
| vegetation | 87.3227 | - |
| terrain | 62.2987 | - |
| sky | 91.6764 | - |
| person | 67.2757 | 49.8798 |
| rider | 50.747 | 27.1253 |
| car | 87.8922 | 81.1109 |
| truck | 21.6999 | 15.2726 |
| bus | 29.0288 | 23.7064 |
| train | 34.7275 | 18.5378 |
| motorcycle | 40.4738 | 19.6388 |
| bicycle | 56.6309 | 38.3812 |
Category results
| Category | IoU | iIoU |
|---|---|---|
| flat | 97.4631 | - |
| nature | 87.1357 | - |
| object | 43.661 | - |
| sky | 91.6764 | - |
| construction | 82.753 | - |
| human | 68.6355 | 51.8936 |
| vehicle | 87.4595 | 80.9352 |