Method Details
Details for method 'SDBNet'
Method overview
| name | SDBNet |
| challenge | pixel-level semantic labeling |
| details | |
| publication | SDBNet: Lightweight Real-time Semantic Segmentation Using Short-term Dense Bottleneck Tanmay Singha, Duc-Son Pham, Aneesh Krishna 2022 International Conference on Digital Image Computing: Techniques and Applications (DICTA) https://ieeexplore.ieee.org/abstract/document/10034634 |
| project page / code | https://github.com/tanmaysingha/SDBNet |
| used Cityscapes data | fine annotations |
| used external data | |
| runtime | n/a |
| subsampling | no |
| submission date | July, 2022 |
| previous submissions |
Average results
| Metric | Value |
|---|---|
| IoU Classes | 70.8289 |
| iIoU Classes | 42.0184 |
| IoU Categories | 87.2469 |
| iIoU Categories | 69.4028 |
Class results
| Class | IoU | iIoU |
|---|---|---|
| road | 97.871 | - |
| sidewalk | 81.3939 | - |
| building | 91.0481 | - |
| wall | 47.3894 | - |
| fence | 47.6228 | - |
| pole | 55.1833 | - |
| traffic light | 64.4752 | - |
| traffic sign | 69.6952 | - |
| vegetation | 92.0342 | - |
| terrain | 68.518 | - |
| sky | 94.373 | - |
| person | 79.4264 | 54.7098 |
| rider | 59.0711 | 32.0714 |
| car | 93.5652 | 85.5243 |
| truck | 53.159 | 21.1581 |
| bus | 70.3832 | 34.1604 |
| train | 62.8693 | 31.8189 |
| motorcycle | 51.5521 | 27.8453 |
| bicycle | 66.1179 | 48.8589 |
Category results
| Category | IoU | iIoU |
|---|---|---|
| flat | 98.2389 | - |
| nature | 91.7248 | - |
| object | 62.7949 | - |
| sky | 94.373 | - |
| construction | 91.1474 | - |
| human | 79.7457 | 55.6685 |
| vehicle | 92.7033 | 83.1371 |