Multi-label Image Recognition on VOC 2007 (test)
96.8mAPResNet-cut + CSRA
Evaluation results
| Method | Links | |||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ResNet-cut + CSRAExtra data (MS-COCO)=true, Backbone=ResNet-101 pretrained with CutMix2021.08 | 96.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SADCLInput resolution=448x448, Reproduced results=true2023.07 | 96.4 | 100 | 99 | 99.5 | 99.1 | 88.9 | 98.8 | 98.7 | 99.6 | 84.2 | 98.4 | 90.1 | 99.4 | 99.6 | 99 | 99.3 | 90.2 | 99.6 | 88.9 | 99.8 | 95.3 | |
| ADD-GCN2020.12 | 96 | 99.8 | 99 | 98.4 | 99 | 86.7 | 98.1 | 98.5 | 98.3 | 85.8 | 98.3 | 88.9 | 98.8 | 99 | 97.4 | 99.2 | 88.3 | 98.7 | 90.7 | 99.5 | 97 | |
| ADDGCNInput resolution=512x512, Reproduced results=true2023.07 | 96 | 99.8 | 99 | 98.4 | 99 | 86.7 | 98.1 | 98.5 | 98.3 | 85.8 | 98.3 | 88.9 | 98.8 | 99 | 97.4 | 99.2 | 88.3 | 98.7 | 90.7 | 99.5 | 97 | |
| ResNet-101 + CSRAExtra data (MS-COCO)=true, Backbone=ResNet-1012021.08 | 96 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ASLExtra data (MS-COCO)=true2021.08 | 95.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MulConInput resolution=448x448, Reproduced results=true2023.07 | 95.6 | 99.8 | 98.3 | 99.3 | 98.6 | 83.3 | 98.4 | 98 | 98.3 | 85.8 | 98.3 | 90.5 | 99.3 | 98.9 | 96.6 | 98.8 | 86.3 | 99.8 | 87.3 | 99.8 | 96.1 | |
| ResNet-cut + CSRAExtra data (MS-COCO)=false, Backbone=ResNet-101 pretrained with CutMix2021.08 | 95.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SSGRL2020.12 | 95 | 99.7 | 98.4 | 98 | 97.6 | 85.7 | 96.2 | 98.2 | 98.8 | 82 | 98.1 | 89.7 | 98.8 | 98.7 | 97 | 99 | 86.9 | 98.1 | 85.8 | 99 | 93.7 | |
| Transformer-based Dual Relation GraphResolution=448x448, Pre-training=ImageNet2021.10 | 95 | 99.9 | 98.9 | 98.4 | 98.7 | 81.9 | 95.8 | 97.8 | 98 | 85.2 | 95.6 | 89.5 | 98.8 | 98.6 | 97.1 | 99.1 | 86.2 | 97.7 | 87.2 | 99.1 | 95.3 | |
| KGGRInput resolution=576x576, Reproduced results=true2023.07 | 95 | 99.3 | 98.6 | 97.9 | 98.4 | 86.2 | 97 | 98 | 99.2 | 82.6 | 98.3 | 87.5 | 99 | 98.9 | 97.4 | 99.1 | 86.9 | 98.2 | 84.1 | 99 | 95 | |
| TDRGInput resolution=448x4482023.07 | 95 | 99.9 | 98.9 | 98.4 | 98.7 | 81.9 | 95.8 | 97.8 | 98 | 85.2 | 95.6 | 89.5 | 98.8 | 98.6 | 97.1 | 99.1 | 86.2 | 97.7 | 87.2 | 99.1 | 95.3 | |
| SSGRLExtra data (MS-COCO)=true, Input image resolution=Larger2021.08 | 95 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GM-MLICInput resolution=448x4482023.07 | 94.7 | 99.4 | 98.7 | 98.5 | 97.6 | 86.3 | 97.1 | 98 | 99.4 | 82.5 | 98.1 | 87.7 | 99.2 | 98.9 | 97.5 | 99.3 | 87 | 98.3 | 86.5 | 99.1 | 94.9 | |
| ResNet-101 + CSRAExtra data (MS-COCO)=false, Backbone=ResNet-1012021.08 | 94.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ASLExtra data (MS-COCO)=false2021.08 | 94.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SSTInput resolution=448x4482023.07 | 94.5 | 99.8 | 98.6 | 98.9 | — | 85.5 | 94.7 | 97.9 | 98.6 | 83 | 96.8 | 85.7 | 98.8 | 98.8 | 98.9 | 99.1 | 85.4 | 96.2 | 84.3 | 99.1 | 95 | |
| DSDLInput resolution=448x4482023.07 | 94.4 | 99.8 | 98.7 | 98.4 | 97.9 | 81.9 | 95.4 | 97.6 | 98.3 | 83.3 | 95 | 88.6 | 98 | 97.9 | 95.8 | 99 | 86.6 | 95.9 | 86.4 | 98.6 | 94.4 | |
| CPCLInput resolution=448x4482023.07 | 94.4 | 99.6 | 98.6 | 98.5 | 98.8 | 81.9 | 95.1 | 97.8 | 98.2 | 83 | 95.5 | 85.5 | 98.4 | 98.5 | 97 | 99 | 86.6 | 97 | 84.9 | 99.1 | 94.3 | |
| P-GCNInput resolution=448x4482023.07 | 94.3 | 99.6 | 98.6 | 98.4 | 98.7 | 81.5 | 94.8 | 97.6 | 98.2 | 83.1 | 96 | 87.1 | 98.3 | 98.5 | 97.3 | 99.1 | 87.3 | 95.5 | 85.4 | 98.9 | 93.6 | |
| ML-GCNBackbone=ResNet-101, Correlation Matrix Type=Re-weighted2019.04 | 94 | 99.5 | 98.5 | 98.6 | 98.1 | 80.8 | 94.6 | 97.2 | 98.2 | 82.3 | 95.7 | 86.4 | 98.2 | 98.4 | 96.7 | 99 | 84.7 | 96.7 | 84.3 | 98.9 | 93.7 | |
| ML-GCN2020.12 | 94 | 99.5 | 98.5 | 98.6 | 98.1 | 80.8 | 94.6 | 97.2 | 98.2 | 82.3 | 95.7 | 86.4 | 98.2 | 98.4 | 96.7 | 99 | 84.7 | 96.7 | 84.3 | 98.9 | 93.7 | |
| ML-GCNResolution=448x448, Pre-training=ImageNet2021.10 | 94 | 99.5 | 98.5 | 98.6 | 98.1 | 80.8 | 94.6 | 97.2 | 98.2 | 82.3 | 95.7 | 86.4 | 98.2 | 98.4 | 96.7 | 99 | 84.7 | 96.7 | 84.3 | 98.9 | 93.7 | |
| ML-GCNInput resolution=448x4482023.07 | 94 | 99.5 | 98.5 | 98.6 | 98.1 | 80.8 | 94.6 | 97.2 | 98.2 | 82.3 | 95.7 | 86.4 | 98.2 | 98.4 | 96.7 | 99 | 84.7 | 96.7 | 84.3 | 98.9 | 93.7 | |
| GCNExtra data (MS-COCO)=false2021.08 | 94 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ResNet-cutExtra data (MS-COCO)=false, Backbone=ResNet-101 pretrained with CutMix2021.08 | 93.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ResNet-101Input resolution=448x448, Reproduced results=true2023.07 | 93.8 | 99 | 98.4 | 97.5 | 96 | 81.4 | 97.3 | 97.3 | 97.1 | 79.6 | 96 | 88.1 | 97.5 | 98.5 | 95.8 | 98.8 | 85.9 | 97.2 | 84.6 | 98.8 | 92 | |
| ADD-GCNResolution=448x448, Pre-training=ImageNet, Implementation=Our implementation2021.10 | 93.6 | 99.7 | 98.5 | 97.6 | 98.4 | 80.6 | 94.1 | 96.6 | 98.1 | 80.4 | 94.9 | 85.7 | 97.9 | 97.9 | 96.4 | 99 | 80.2 | 97.3 | 85.3 | 98.9 | 94.1 | |
| SSGRLResolution=448x448, Pre-training=ImageNet2021.10 | 93.4 | 99.5 | 97.1 | 97.6 | 97.8 | 82.6 | 94.8 | 96.7 | 98.1 | 78 | 97 | 85.6 | 97.8 | 98.3 | 96.1 | 98.1 | 84.9 | 96.5 | 79.8 | 98.4 | 92.8 | |
| SSGRLInput resolution=576x576, Reproduced results=true2023.07 | 93.4 | 99.5 | 97.1 | 97.6 | 97.8 | 82.6 | 94.8 | 96.7 | 98.1 | 78 | 97 | 85.6 | 97.8 | 98.3 | 95.7 | 98.1 | 84.9 | 96.5 | 79.8 | 98.4 | 92.8 | |
| SSGRLExtra data (MS-COCO)=false, Input image resolution=Larger2021.08 | 93.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ML-GCNBackbone=ResNet-101, Correlation Matrix Type=Binary2019.04 | 93.1 | 99.6 | 98.3 | 97.9 | 97.6 | 78.2 | 92.3 | 97.4 | 97.4 | 79.2 | 94.4 | 86.5 | 97.4 | 97.9 | 97.1 | 98.7 | 84.6 | 95.3 | 83 | 98.6 | 90.4 | |
| ResNet-101Resolution=448x448, Pre-training=ImageNet, Implementation=Our implementation2021.10 | 92.9 | 99.8 | 98.3 | 98 | 98 | 79.5 | 93.2 | 96.8 | 97.7 | 79.9 | 91 | 86.6 | 98.2 | 97.8 | 96.1 | 98.8 | 79.4 | 94.6 | 82.9 | 99.1 | 92.1 | |
| ResNet-101Extra data (MS-COCO)=false, Backbone=ResNet-1012021.08 | 92.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ML-GCNBackbone=VGG, Correlation Matrix Type=Re-weighted2019.04 | 92.8 | 99.4 | 97.4 | 98 | 97 | 77.9 | 92.4 | 96.8 | 97.8 | 80.8 | 93.4 | 87.2 | 98 | 97.3 | 95.8 | 98.8 | 79.4 | 95.3 | 82.2 | 99.1 | 91.4 | |
| RCP2020.12 | 92.5 | 99.3 | 97.6 | 98 | 96.4 | 79.3 | 93.8 | 96.6 | 97.1 | 78 | 88.7 | 87.1 | 97.1 | 96.3 | 95.4 | 99.1 | 82.1 | 93.6 | 82.2 | 98.4 | 92.8 | |
| RCPExtra data (MS-COCO)=false2021.08 | 92.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Atten-Reinforce2019.04 | 92 | 98.6 | 97.1 | 97.1 | 95.5 | 75.6 | 92.8 | 96.8 | 97.3 | 78.3 | 92.2 | 87.6 | 96.9 | 96.5 | 93.6 | 98.5 | 81.6 | 93.1 | 83.2 | 98.5 | 89.3 | |
| FeV+LV2020.12 | 92 | 98.2 | 96.9 | 97.1 | 95.8 | 74.3 | 94.2 | 96.7 | 96.7 | 76.7 | 90.5 | 88 | 96.9 | 97.7 | 95.9 | 98.6 | 78.5 | 93.6 | 82.4 | 98.4 | 90.4 | |
| RARL2020.12 | 92 | 98.6 | 97.1 | 97.1 | 95.5 | 75.6 | 92.8 | 96.8 | 97.3 | 78.3 | 92.2 | 87.6 | 96.9 | 96.5 | 93.6 | 98.5 | 81.6 | 93.1 | 83.2 | 98.5 | 89.3 | |
| Atten-ReinforceResolution=448x448, Pre-training=ImageNet2021.10 | 92 | 98.6 | 97.1 | 97.1 | 95.5 | 75.6 | 92.8 | 96.8 | 97.3 | 78.3 | 92.2 | 87.6 | 96.9 | 96.5 | 93.6 | 98.5 | 81.6 | 93.1 | 83.2 | 98.5 | 89.3 | |
| RNN-Attention2019.04 | 91.9 | 98.6 | 97.4 | 96.3 | 96.2 | 75.2 | 92.4 | 96.5 | 97.1 | 76.5 | 92 | 87.7 | 96.8 | 97.5 | 93.8 | 98.5 | 81.6 | 93.7 | 82.8 | 98.6 | 89.3 | |
| RDAR2020.12 | 91.9 | 98.6 | 97.4 | 96.3 | 96.2 | 75.2 | 92.4 | 96.5 | 97.1 | 76.5 | 92 | 87.7 | 96.8 | 97.5 | 93.8 | 98.5 | 81.6 | 93.7 | 82.8 | 98.6 | 89.3 | |
| RNN-AttentionResolution=448x448, Pre-training=ImageNet2021.10 | 91.9 | 98.6 | 97.4 | 96.3 | 96.2 | 75.2 | 92.4 | 96.5 | 97.1 | 76.5 | 92 | 87.7 | 96.8 | 97.5 | 93.8 | 98.5 | 81.6 | 93.7 | 82.8 | 98.6 | 89.3 | |
| RNN-AttInput resolution=448x4482023.07 | 91.9 | 98.6 | 97.4 | 96.3 | 96.2 | 75.2 | 92.4 | 96.5 | 97.1 | 76.5 | 92 | 87.7 | 96.8 | 97.5 | 94.8 | 96.7 | 81.6 | 93.7 | 82.8 | 98.6 | 89.3 | |
| ML-GCNBackbone=VGG, Correlation Matrix Type=Binary2019.04 | 91.1 | 98.3 | 97.1 | 96.1 | 96.7 | 75 | 91.4 | 95.8 | 95.4 | 76.7 | 92.1 | 85.1 | 96.7 | 96 | 95.3 | 97.8 | 77.4 | 93.1 | 79.7 | 97.9 | 89.3 | |
| HCP2019.04 | 90.9 | 98.6 | 97.1 | 98 | 95.6 | 75.3 | 94.7 | 95.8 | 97.3 | 73.1 | 90.2 | 80 | 97.3 | 96.1 | 94.9 | 96.3 | 78.3 | 94.7 | 76.2 | 97.9 | 91.5 | |
| HCP2020.12 | 90.9 | 98.6 | 97.1 | 98 | 95.6 | 75.3 | 94.7 | 95.8 | 97.3 | 73.1 | 90.2 | 80 | 97.3 | 96.1 | 94.9 | 96.3 | 78.3 | 94.7 | 76.2 | 97.9 | 91.5 | |
| ResNet-101backbone=ResNet-1012020.12 | 90.8 | 99.1 | 97.3 | 96.2 | 94.7 | 68.3 | 92.9 | 95.9 | 94.6 | 77.9 | 89.9 | 85.1 | 94.7 | 96.8 | 94.3 | 98.1 | 80.8 | 93.1 | 79.1 | 98.2 | 91.1 | |
| FeV+LV2019.04 | 90.6 | 97.9 | 97 | 96.6 | 94.6 | 73.6 | 93.9 | 96.5 | 95.5 | 73.7 | 90.3 | 82.8 | 95.4 | 97.7 | 95.9 | 98.6 | 77.6 | 88.7 | 78 | 98.3 | 89 | |
| Fev+LvResolution=448x448, Pre-training=ImageNet2021.10 | 90.6 | 97.9 | 97 | 96.6 | 94.6 | 73.6 | 93.9 | 96.5 | 95.5 | 73.7 | 90.3 | 82.8 | 95.4 | 97.7 | 95.9 | 98.6 | 77.6 | 88.7 | 78 | 98.3 | 89 | |
| ResNet-1012019.04 | 89.9 | 99.5 | 97.7 | 97.8 | 96.4 | 65.7 | 91.8 | 96.1 | 97.6 | 74.2 | 80.9 | 85 | 98.4 | 96.5 | 95.9 | 98.4 | 70.1 | 88.3 | 80.2 | 98.9 | 89.2 | |
| VeryDeep2019.04 | 89.7 | 98.9 | 95 | 96.8 | 95.4 | 69.7 | 90.4 | 93.5 | 96 | 74.2 | 86.6 | 87.8 | 96 | 96.3 | 93.1 | 97.2 | 70 | 92.1 | 80.3 | 98.1 | 87 | |
| Very Deep2020.12 | 89.7 | 98.9 | 95 | 96.8 | 95.4 | 69.7 | 90.4 | 93.5 | 96 | 74.2 | 86.6 | 87.8 | 96 | 96.3 | 93.1 | 97.2 | 70 | 92.1 | 80.3 | 98.1 | 87 | |
| RLSD2019.04 | 88.5 | 96.4 | 92.7 | 93.8 | 94.1 | 71.2 | 92.5 | 94.2 | 95.7 | 74.3 | 90 | 74.2 | 95.4 | 96.2 | 92.1 | 97.9 | 66.9 | 93.5 | 73.7 | 97.5 | 87.6 | |
| RLSD2020.12 | 88.5 | 96.4 | 92.7 | 93.8 | 94.1 | 71.2 | 92.5 | 94.2 | 95.7 | 74.3 | 90 | 74.2 | 95.4 | 96.2 | 92.1 | 97.9 | 66.9 | 93.5 | 73.7 | 97.5 | 87.6 | |
| RMIC2020.12 | 84.5 | 97.1 | 91.3 | 94.2 | 57.1 | 86.7 | 90.7 | 93.1 | 63.3 | 83.3 | 76.4 | 92.8 | 94.4 | 91.6 | 95.1 | 92.3 | 59.7 | 86 | 69.5 | 96.4 | 79 | |
| CNN-RNN2019.04 | 84 | 96.7 | 83.1 | 94.2 | 92.8 | 61.2 | 82.1 | 89.1 | 94.2 | 64.2 | 83.6 | 70 | 92.4 | 91.7 | 84.2 | 93.7 | 59.8 | 93.2 | 75.3 | 99.7 | 78.6 | |
| CNN-RNN2020.12 | 84 | 96.7 | 83.1 | 94.2 | 92.8 | 61.2 | 82.1 | 89.1 | 94.2 | 64.2 | 83.6 | 70 | 92.4 | 91.7 | 84.2 | 93.7 | 59.8 | 93.2 | 75.3 | 99.7 | 78.6 | |
| CNN-RNNResolution=448x448, Pre-training=ImageNet2021.10 | 84 | 96.7 | 83.1 | 94.2 | 92.8 | 61.2 | 82.1 | 89.1 | 94.2 | 64.2 | 83.6 | 70 | 92.4 | 91.7 | 84.2 | 93.7 | 59.8 | 93.2 | 75.3 | 99.7 | 78.6 | |
| CNN-RNNInput resolution=448x4482023.07 | 84 | 96.7 | 83.1 | 94.2 | 92.8 | 61.2 | 82.1 | 89.1 | 94.2 | 64.2 | 83.6 | 70 | 92.4 | 91.7 | 84.2 | 93.7 | 59.8 | 93.2 | 75.3 | 99.7 | 78.6 |