Metric: R@1 (higher is better)
| # | Model↕ | R@1▼ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | CGD (MG/SG) | 94.8 | No | Combination of Multiple Global Descriptors for I... | 2019-03-26 | Code |
| 2 | ProxyNCA++ | 90.1 | No | ProxyNCA++: Revisiting and Revitalizing Proxy Ne... | 2020-04-02 | Code |
| 3 | NormSoftmax2048 (ResNet-50) | 89.3 | No | Classification is a Strong Baseline for Deep Met... | 2018-11-30 | Code |
| 4 | MES-Loss | 87.89 | No | - | - | - |
| 5 | Margin | 86.9 | No | Sampling Matters in Deep Embedding Learning | 2017-06-23 | Code |
| 6 | MS512 | 84.1 | No | Multi-Similarity Loss with General Pair Weightin... | 2019-04-14 | Code |
| 7 | EPSHN512 | 82.7 | No | Improved Embeddings with Easy Positive Triplet M... | 2019-04-08 | Code |
| 8 | HTL | 81.4 | No | Deep Metric Learning with Hierarchical Triplet L... | 2018-10-16 | - |