Method Details
Details for method 'TKCN'
Method overview
| name | TKCN |
| challenge | pixel-level semantic labeling |
| details | |
| publication | Tree-structured Kronecker Convolutional Network for Semantic Segmentation Tianyi Wu, Sheng Tang, Rui Zhang, Juan Cao, Jintao Li https://arxiv.org/abs/1812.04945 |
| project page / code | https://github.com/wutianyiRosun/TKCN |
| used Cityscapes data | fine annotations |
| used external data | ImageNet |
| runtime | n/a |
| subsampling | no |
| submission date | December, 2018 |
| previous submissions | 1, 2 |
Average results
| Metric | Value |
|---|---|
| IoU Classes | 79.4531 |
| iIoU Classes | 61.2833 |
| IoU Categories | 91.0718 |
| iIoU Categories | 81.5192 |
Class results
| Class | IoU | iIoU |
|---|---|---|
| road | 98.416 | - |
| sidewalk | 85.7831 | - |
| building | 92.9526 | - |
| wall | 51.6689 | - |
| fence | 61.6981 | - |
| pole | 67.5958 | - |
| traffic light | 75.7606 | - |
| traffic sign | 80.002 | - |
| vegetation | 93.6263 | - |
| terrain | 72.7388 | - |
| sky | 95.4384 | - |
| person | 86.8596 | 72.4314 |
| rider | 70.8889 | 52.8531 |
| car | 95.8876 | 90.9401 |
| truck | 64.4598 | 43.5235 |
| bus | 86.9121 | 58.265 |
| train | 81.7741 | 54.4555 |
| motorcycle | 69.5908 | 51.1191 |
| bicycle | 77.5552 | 66.6789 |
Category results
| Category | IoU | iIoU |
|---|---|---|
| flat | 98.5843 | - |
| nature | 93.3265 | - |
| object | 74.0881 | - |
| sky | 95.4384 | - |
| construction | 93.5448 | - |
| human | 87.3175 | 73.5633 |
| vehicle | 95.2029 | 89.4751 |