Visual Place Recognition On Nordland
Metrics
Recall@1
Recall@5
Results
Performance results of various models on this benchmark
Model Name | Recall@1 | Recall@5 | Paper Title | Repository |
|---|---|---|---|---|
| SelaVPR | 86.6 | 94.0 | Towards Seamless Adaptation of Pre-trained Models for Visual Place Recognition | |
| BoQ | 90.6 | 96.0 | BoQ: A Place is Worth a Bag of Learnable Queries | |
| MixVPR | 76.0 | 89.2 | MixVPR: Feature Mixing for Visual Place Recognition | |
| DINOv2 SALAD (1-frame thr.) | 85.2 | 98.5 | Optimal Transport Aggregation for Visual Place Recognition | |
| Patch-NetVLAD | 58.4 | 74.6 | Patch-NetVLAD: Multi-Scale Fusion of Locally-Global Descriptors for Place Recognition | |
| NetVLAD (with GPM) | 44.9 | 50.2 | Global Proxy-based Hard Mining for Visual Place Recognition | |
| EffoVPR | 95.0 | 98.6 | EffoVPR: Effective Foundation Model Utilization for Visual Place Recognition | - |
| BoQ (ResNet-50) | 83.1 | - | BoQ: A Place is Worth a Bag of Learnable Queries | |
| Conv-AP | 38.5 | 53.9 | GSV-Cities: Toward Appropriate Supervised Visual Place Recognition | |
| HEAPUtil | 38.2 | 54.8 | A Hierarchical Dual Model of Environment- and Place-Specific Utility for Visual Place Recognition |
0 of 10 row(s) selected.