Method Details
Details for method 'Adelaide'
Method overview
| name | Adelaide |
| challenge | pixel-level semantic labeling |
| details | Trained on a pre-release version of the dataset |
| publication | Efficient Piecewise Training of Deep Structured Models for Semantic Segmentation G. Lin, C. Shen, I. Reid, and A. van den Hengel arXiv preprint 2015 http://arxiv.org/pdf/1504.01013v2 |
| project page / code | |
| used Cityscapes data | fine annotations |
| used external data | ImageNet |
| runtime | 35 s |
| subsampling | no |
| submission date | April, 2016 |
| previous submissions |
Average results
| Metric | Value |
|---|---|
| IoU Classes | 66.399 |
| iIoU Classes | 46.7273 |
| IoU Categories | 82.7603 |
| iIoU Categories | 67.4338 |
Class results
| Class | IoU | iIoU |
|---|---|---|
| road | 97.2711 | - |
| sidewalk | 78.5006 | - |
| building | 88.3844 | - |
| wall | 44.4675 | - |
| fence | 48.2643 | - |
| pole | 34.096 | - |
| traffic light | 55.4515 | - |
| traffic sign | 61.6525 | - |
| vegetation | 90.0663 | - |
| terrain | 69.503 | - |
| sky | 92.2424 | - |
| person | 72.4868 | 56.2414 |
| rider | 52.283 | 38.0152 |
| car | 90.9574 | 77.1012 |
| truck | 54.6381 | 34.0072 |
| bus | 61.6087 | 47.0492 |
| train | 51.5948 | 33.4023 |
| motorcycle | 55.0289 | 38.1337 |
| bicycle | 63.0828 | 49.8688 |
Category results
| Category | IoU | iIoU |
|---|---|---|
| flat | 97.8021 | - |
| nature | 89.7339 | - |
| object | 48.1585 | - |
| sky | 92.2424 | - |
| construction | 88.6563 | - |
| human | 73.1264 | 58.2034 |
| vehicle | 89.6025 | 76.6643 |