Method Details
Details for method 'iFlytek-ZBGKRD-fcos3d-depth-norm'
Method overview
| name | iFlytek-ZBGKRD-fcos3d-depth-norm |
| challenge | 3D vehicle detection |
| details | |
| publication | Anonymous |
| project page / code | |
| used Cityscapes data | fine annotations, 3D bounding boxes |
| used external data | |
| runtime | n/a |
| subsampling | no |
| submission date | August, 2021 |
| previous submissions |
Average results
| Metric | Value |
|---|---|
| Detection Score | 42.936 |
| AP | 47.5769 |
| Center Distance | 96.5855 |
| Size Similarity | 80.3611 |
| Orientation Similarity Yaw | 80.371 |
| Orientation Similarity Pitch Roll | 99.965 |
Class results
| Class | Detection Score |
|---|---|
| car | 75.8405 |
| truck | 33.2838 |
| bus | 41.7388 |
| train | 23.6409 |
| motorcycle | 39.6093 |
| bicycle | 43.5026 |
Results json
This json can be visualized using the tool csPlot3dDetectionResults,
which is part of the cityscapesScripts found on
Github.