Visual Place Recognition On Pittsburgh 250K
Metrics
Recall@1
Recall@10
Recall@5
Results
Performance results of various models on this benchmark
Model Name | Recall@1 | Recall@10 | Recall@5 | Paper Title | Repository |
|---|---|---|---|---|---|
| SelaVPR | 95.7 | 98.8 | 99.2 | Towards Seamless Adaptation of Pre-trained Models for Visual Place Recognition | |
| ProGEO | 92.2 | - | 97.7 | ProGEO: Generating Prompts through Image-Text Contrastive Learning for Visual Geo-localization | |
| BoQ | 96.6 | 99.5 | 99.1 | BoQ: A Place is Worth a Bag of Learnable Queries | |
| NetVLAD (with GPM) | 91.5 | 98.1 | 97.2 | Global Proxy-based Hard Mining for Visual Place Recognition | |
| BoQ (ResNet-50) | 95 | 99.1 | 98.5 | BoQ: A Place is Worth a Bag of Learnable Queries | |
| DINOv2 SALAD | 95.1 | 99.1 | 98.5 | Optimal Transport Aggregation for Visual Place Recognition | |
| MixVPR | 94.6 | 99.0 | 98.3 | MixVPR: Feature Mixing for Visual Place Recognition | |
| CosPlace | 91.5 | 97.9 | 96.9 | Rethinking Visual Geo-localization for Large-Scale Applications | |
| EigenPlaces | 94.1 | - | - | EigenPlaces: Training Viewpoint Robust Models for Visual Place Recognition | |
| Conv-AP | 92.4 | 98.6 | 97.6 | GSV-Cities: Toward Appropriate Supervised Visual Place Recognition |
0 of 10 row(s) selected.