Method Details
Details for method 'ESANet RGB-D'
Method overview
| name | ESANet RGB-D |
| challenge | pixel-level semantic labeling |
| details | Efficient RGB-D Semantic Segmentation for Indoor Scene Analysis. ESANet-R34-NBt1D using RGB-D data. |
| publication | Efficient RGB-D Semantic Segmentation for Indoor Scene Analysis Daniel Seichter, Mona Köhler, Benjamin Lewandowski, Tim Wengefeld and Horst-Michael Gross |
| project page / code | https://github.com/TUI-NICR/ESANet |
| used Cityscapes data | fine annotations, stereo |
| used external data | ImageNet |
| runtime | 0.1613 s NVIDIA Jetson AGX Xavier (Jetpack 4.4, TensorRT 7.1, Float16) |
| subsampling | no |
| submission date | November, 2020 |
| previous submissions |
Average results
| Metric | Value |
|---|---|
| IoU Classes | 78.4156 |
| iIoU Classes | 56.4287 |
| IoU Categories | 91.3277 |
| iIoU Categories | 78.9692 |
Class results
| Class | IoU | iIoU |
|---|---|---|
| road | 98.6801 | - |
| sidewalk | 87.1258 | - |
| building | 93.3235 | - |
| wall | 49.8248 | - |
| fence | 60.1854 | - |
| pole | 69.0855 | - |
| traffic light | 76.0786 | - |
| traffic sign | 79.4331 | - |
| vegetation | 93.5753 | - |
| terrain | 72.6651 | - |
| sky | 95.8992 | - |
| person | 87.2127 | 68.3665 |
| rider | 71.2892 | 48.7278 |
| car | 96.0807 | 90.7017 |
| truck | 66.2357 | 36.9913 |
| bus | 78.4268 | 49.8291 |
| train | 71.5065 | 48.718 |
| motorcycle | 67.102 | 46.2371 |
| bicycle | 76.1663 | 61.8579 |
Category results
| Category | IoU | iIoU |
|---|---|---|
| flat | 98.7184 | - |
| nature | 93.3121 | - |
| object | 75.0653 | - |
| sky | 95.8992 | - |
| construction | 93.7242 | - |
| human | 87.1026 | 69.2849 |
| vehicle | 95.4722 | 88.6535 |
