Visual Place Recognition On Baidu Mall
评估指标
Recall@1
评测结果
各个模型在此基准测试上的表现结果
模型名称 | Recall@1 | Paper Title | Repository |
|---|---|---|---|
| DINO | 48.30 | Emerging Properties in Self-Supervised Vision Transformers | |
| MixVPR | 64.44 | MixVPR: Feature Mixing for Visual Place Recognition | |
| SegVLAD-PreT (M) | 80.4 | Revisit Anything: Visual Place Recognition via Image Segment Retrieval | |
| CLIP | 56.02 | AnyLoc: Towards Universal Visual Place Recognition | |
| DINOv2 | 49.21 | DINOv2: Learning Robust Visual Features without Supervision | |
| NetVLAD | 53.10 | NetVLAD: CNN architecture for weakly supervised place recognition | |
| AnyLoc-VLAD-DINOv2 | 75.22 | AnyLoc: Towards Universal Visual Place Recognition | |
| CosPlace | 41.62 | Rethinking Visual Geo-localization for Large-Scale Applications |
0 of 8 row(s) selected.