Learning With Coarse Labels On Stanford
Metrics
Recall@1
Recall@10
Recall@2
Recall@5
Results
Performance results of various models on this benchmark
| Paper Title | |||||
|---|---|---|---|---|---|
| MaskCon | 74.05 | 87.96 | 78.97 | 84.48 | MaskCon: Masked Contrastive Learning for Coarse-Labelled Dataset |
| Grafit | 74.02 | 87.91 | 78.82 | 84.13 | Grafit: Learning fine-grained image representations with coarse labels |
0 of 2 row(s) selected.