Multi-label Classification on Open Images (test)
85mAPBiAM
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| BiAMBackbone=VGG-192021.08 | 85 | 0.204 | 0.173 | — | — | — | — | |
| CLF (Our Approach)2021.01 | 76.9 | 0.37 | 0.327 | 0.34 | 0.407 | 0.233 | 0.549 | |
| Logistic2021.01 | 75.1 | 0.133 | 0.118 | 0.121 | 0.147 | 0.084 | 0.202 | |
| WSABIE2021.01 | 71.7 | 0.022 | 0.022 | 0.015 | 0.037 | 0.015 | 0.037 | |
| WARP2021.01 | 69.9 | 0.077 | 0.074 | 0.071 | 0.085 | 0.053 | 0.126 | |
| LESA2021.01 | 69.3 | 0.178 | 0.145 | 0.162 | 0.196 | 0.103 | 0.247 | |
| Fast0Tag2021.01 | 69 | 0.162 | 0.131 | 0.149 | 0.179 | 0.093 | 0.223 | |
| One Attention per Cluster2021.01 | 68.5 | 0.163 | 0.13 | 0.149 | 0.179 | 0.092 | 0.22 | |
| CNN-RNN2021.01 | 62.3 | 0.096 | 0.105 | 0.087 | 0.105 | 0.054 | 0.131 | |
| LogisticBackbone=VGG-192021.08 | 49.4 | 0.133 | 0.118 | — | — | — | — | |
| WSABIEBackbone=VGG-192021.08 | 47.2 | 0.022 | 0.022 | — | — | — | — | |
| WARPBackbone=VGG-192021.08 | 46 | 0.077 | 0.074 | — | — | — | — | |
| LESABackbone=VGG-192021.08 | 45.6 | 0.178 | 0.145 | — | — | — | — | |
| Fast0TagBackbone=VGG-192021.08 | 45.4 | 0.162 | 0.131 | — | — | — | — | |
| One Attention per ClusterBackbone=VGG-192021.08 | 45.1 | 0.163 | 0.13 | — | — | — | — | |
| CNN-RNNBackbone=VGG-192021.08 | 41 | 0.096 | 0.105 | — | — | — | — |