Omnnidirectional Stereo Depth Estimation On
评估指标
Depth-LRCE
Depth-MAE
Depth-MARE
Depth-RMSE
Disp-MAE
Disp-MARE
Disp-RMSE
评测结果
各个模型在此基准测试上的表现结果
模型名称 | Depth-LRCE | Depth-MAE | Depth-MARE | Depth-RMSE | Disp-MAE | Disp-MARE | Disp-RMSE | Paper Title | Repository |
|---|---|---|---|---|---|---|---|---|---|
| 360SD-Net | 0.904 | 2.112 | 0.152 | 5.077 | 0.224 | 0.191 | 0.419 | 360SD-Net: 360° Stereo Depth Estimation with Learnable Cost Volume | - |
| 360-IGEV-Stereo | 0.388 | 1.720 | 0.130 | 4.297 | 0.188 | 0.146 | 0.404 | Helvipad: A Real-World Dataset for Omnidirectional Stereo Depth Estimation | |
| IGEV-Stereo | 1.203 | 1.860 | 0.146 | 4.474 | 0.225 | 0.172 | 0.423 | Iterative Geometry Encoding Volume for Stereo Matching | |
| PSMNet | 1.809 | 2.509 | 0.176 | 5.673 | 0.286 | 0.248 | 0.496 | Pyramid Stereo Matching Network |
0 of 4 row(s) selected.