Visual Localization on Extended CMU Seasons (Park)
92.1Recall @ (0.25m, 2°)DeViLoc
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| DeViLocMethod Group=SD2024.02 | 92.1 | 95.1 | 96.3 | |
| D2NetMethod Group=S2024.02 | 89.2 | 93.2 | 95 | |
| HLoc+PixLocMethod Group=S2024.02 | 87 | 89.5 | 91.6 | |
| HLoc[SP+SG]Method Group=S2024.02 | 85.7 | 89 | 91.6 | |
| PFSLBackbone=FGSN, Clusters=200, Training dataset=CMU2019.08 | 64.8 | 81.5 | 89.3 | |
| PixLocMethod Group=S2024.02 | 61 | 62.5 | 69.4 | |
| HF-Net2019.08 | 57.4 | 64.4 | 80.4 | |
| HLoc[SP]Method Group=S2024.02 | 57.4 | 64.4 | 80.4 | |
| Asymmetric Hypercolumn Matching2019.08 | 54.3 | 71.6 | 84.1 | |
| SSMCBackbone=FGSN, Clusters=100, Training dataset=CMU2019.08 | 51.4 | 57.6 | 65.5 | |
| ASMethod Group=S2024.02 | 45.5 | 51.6 | 62 | |
| PFSL2019.08 | 39 | 61.2 | 95.6 | |
| GSMCBackbone=FGSN, Clusters=200, Training dataset=CMU2019.08 | 38.9 | 43.4 | 50 | |
| GSMC2019.08 | 37.8 | 42 | 49.3 | |
| City Scale Localization2019.08 | 34.5 | 37 | 42.2 | |
| DenseVLAD2019.08 | 5.2 | 19.1 | 62 | |
| NetVLAD2019.08 | 2.6 | 10.4 | 55.9 |