Rotation estimation on CAST drone dataset real-world tracking scenario 1897 RGB-D images
0.456Mean Runtime (ms)SCF
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| SCFRelaxation=SO(3), Keypoint detections (N)=7, Shape library (K)=10, lambda=0.12025.09 | 0.456 | 0.774 | 9.4 | 7.1 | |
| SCF*Relaxation=O(3), Keypoint detections (N)=7, Shape library (K)=10, lambda=0.12025.09 | 0.606 | 1.029 | 9.4 | 7.1 | |
| G-NKeypoint detections (N)=7, Shape library (K)=10, lambda=0.12025.09 | 1.817 | 3.48 | 9.5 | 7.1 | |
| L-MKeypoint detections (N)=7, Shape library (K)=10, lambda=0.12025.09 | 5.2 | 12.701 | 9.5 | 7.1 | |
| ManoptKeypoint detections (N)=7, Shape library (K)=10, lambda=0.12025.09 | 6.062 | 11.783 | 9.5 | 7.1 | |
| PACE*Keypoint detections (N)=7, Shape library (K)=10, lambda=0.12025.09 | 10.826 | 21.514 | 9.5 | 7.1 |