Image Captioning on MS-COCO online (test)
42.5BLEU-4 (c5)GRIT
Evaluation Results
| Method | Links | |||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GRITEnsemble=true, VL pretraining=false2022.07 | 42.5 | 76.8 | 30.9 | 41 | 61.2 | 77.1 | 141.3 | 143.8 | 84.1 | 97.6 | 69.4 | 93.5 | 54.9 | 86.3 | — | — | — | — | — | — | — | |
| GRITEnsemble=true, VL pretraining=false2026.05 | 42.5 | 76.8 | 30.9 | 41 | 61.2 | 77.1 | 141.3 | 143.8 | 84.1 | 97.6 | 69.4 | 93.5 | 54.9 | 86.3 | — | — | — | — | — | — | — | |
| COS-Netsplit=official test, server=online test server2022.06 | 42 | 74.7 | 30.4 | 40.1 | 60.6 | 76.4 | 136.7 | 138.3 | 83.3 | 96.8 | 68.6 | 92.3 | 54.2 | 84.5 | — | — | — | — | — | — | — | |
| HAAVEnsemble=4 models2023.05 | 41.7 | 76.1 | 30.2 | 39.9 | 60.4 | 75.8 | 139.1 | 142.3 | 84 | 97.6 | 69.1 | 93.3 | 54.3 | 85.8 | — | — | — | — | — | — | — | |
| GRITEnsemble=false, VL pretraining=false2022.07 | 41.5 | 75.6 | 30.3 | 40.2 | 60.2 | 75.9 | 138.3 | 141.8 | 83.7 | 97.4 | 68.5 | 92.8 | 53.9 | 85.3 | — | — | — | — | — | — | — | |
| GRITEnsemble=false, VL pretraining=false2026.05 | 41.5 | 75.6 | 30.3 | 40.2 | 60.2 | 75.9 | 138.3 | 141.8 | 83.7 | 97.4 | 68.5 | 92.8 | 53.9 | 85.3 | — | — | — | — | — | — | — | |
| MD-SAN2022.01 | 40.7 | 73.7 | 29.8 | 39.4 | 59.8 | 75 | 133.4 | 135.4 | 82.4 | 96.5 | 67.4 | 91.6 | 52.8 | 83.6 | — | — | — | — | — | — | — | |
| DLCTEnsemble=true, VL pretraining=false2022.07 | 40.6 | 74 | 29.8 | 39.6 | 59.8 | 75.3 | 133.3 | 135.4 | 82.4 | 96.6 | 67.4 | 91.7 | 52.8 | 83.8 | — | — | — | — | — | — | — | |
| DLCTEnsemble=4 models2023.05 | 40.6 | 74 | 29.8 | 39.6 | 59.8 | 75.3 | 133.3 | 135.4 | 82.4 | 96.6 | 67.4 | 91.7 | 52.8 | 83.8 | — | — | — | — | — | — | — | |
| I^2OA2022.01 | 40.6 | 74 | 29.9 | 39.5 | 59.8 | 75.3 | 133.7 | 135.6 | 82.4 | 96.6 | 67.4 | 91.6 | 52.8 | 83.8 | — | — | — | — | — | — | — | |
| DLCTEnsemble=true, VL pretraining=false2026.05 | 40.6 | 74 | 29.8 | 39.6 | 59.8 | 75.3 | 133.3 | 135.4 | 82.4 | 96.6 | 67.4 | 91.7 | 52.8 | 83.8 | — | — | — | — | — | — | — | |
| CBTrans2022.01 | 40.5 | 73.5 | 30 | 39.6 | 59.6 | 75 | 135 | 138.6 | 82.3 | 96.5 | 67.5 | 91.8 | 52.8 | 83.7 | — | — | — | — | — | — | — | |
| OSCAR+ w/ VinVL2021.01 | 40.4 | 74.9 | 30.6 | 40.8 | 60.4 | 76.8 | 134.7 | 138.7 | 81.9 | 96.9 | 66.9 | 92.4 | 52.6 | 84.7 | — | — | — | — | — | — | — | |
| VinVLlargeEnsemble=false, VL pretraining=true2022.07 | 40.4 | 74.9 | 30.6 | 40.8 | 60.4 | 76.8 | 134.7 | 138.7 | 81.9 | 96.9 | 66.9 | 92.4 | 52.6 | 84.7 | — | — | — | — | — | — | — | |
| VinVLlargeEnsemble=false, VL pretraining=true2026.05 | 40.4 | 74.9 | 30.6 | 40.8 | 60.4 | 76.8 | 134.7 | 138.7 | 81.9 | 96.9 | 66.9 | 92.4 | 52.6 | 84.7 | — | — | — | — | — | — | — | |
| X-Transformer2021.01 | 40.3 | 72.4 | 29.6 | 39.2 | 59.5 | 75 | 131.1 | 133.5 | 81.9 | 95.7 | 66.9 | 90.5 | 52.4 | 82.5 | — | — | — | — | — | — | — | |
| X-LANEnsemble=true, VL pretraining=false2022.07 | 40.3 | 72.4 | 29.6 | 39.2 | 59.5 | 75 | 131.1 | 133.5 | 81.9 | 95.7 | 66.9 | 90.5 | 52.4 | 82.5 | — | — | — | — | — | — | — | |
| X-LANEnsemble=4 models2023.05 | 40.3 | 72.4 | 29.6 | 39.2 | 59.5 | 75 | 131.1 | 133.5 | 81.9 | 95.7 | 66.9 | 90.5 | 52.4 | 82.5 | — | — | — | — | — | — | — | |
| X-Transformer2022.01 | 40.3 | 72.4 | 29.6 | 39.2 | 59.5 | 75 | 131.1 | 133.5 | 81.9 | 95.7 | 66.9 | 90.5 | 52.4 | 82.5 | — | — | — | — | — | — | — | |
| X-LANEnsemble=true, VL pretraining=false2026.05 | 40.3 | 72.4 | 29.6 | 39.2 | 59.5 | 75 | 131.1 | 133.5 | 81.9 | 95.7 | 66.9 | 90.5 | 52.4 | 82.5 | — | — | — | — | — | — | — | |
| RSTNet2022.01 | 40 | 73.1 | 29.6 | 39.1 | 59.5 | 74.6 | 131.9 | 134 | 82.1 | 96.4 | 67 | 91.3 | 52.2 | 83 | — | — | — | — | — | — | — | |
| X-Transformersplit=official test, server=online test server2022.06 | 39.9 | 71.8 | 29.5 | 39 | 59.3 | 74.9 | 129.3 | 131.4 | 81.3 | 95.4 | 66.3 | 90 | 51.9 | 81.7 | — | — | — | — | — | — | — | |
| M2 Transformersplit=official test, server=online test server2022.06 | 39.7 | 72.8 | 29.4 | 39 | 59.2 | 74.8 | 129.3 | 132.1 | 81.6 | 96 | 66.4 | 90.8 | 51.8 | 82.7 | — | — | — | — | — | — | — | |
| M²Trans.Ensemble=true, VL pretraining=false2022.07 | 39.7 | 72.8 | 29.4 | 39 | 59.2 | 74.8 | 129.3 | 132.1 | 81.6 | 96 | 66.4 | 90.8 | 51.8 | 82.7 | — | — | — | — | — | — | — | |
| M²Ensemble=4 models2023.05 | 39.7 | 72.8 | 29.4 | 39 | 59.2 | 74.8 | 129.3 | 132.1 | 81.6 | 96 | 66.4 | 90.8 | 51.8 | 82.7 | — | — | — | — | — | — | — | |
| M^2 T2022.01 | 39.7 | 72.8 | 29.4 | 39 | 59.2 | 74.8 | 129.3 | 132.1 | 81.6 | 96 | 66.4 | 90.8 | 51.8 | 82.7 | — | — | — | — | — | — | — | |
| GET2022.01 | 39.7 | 72.9 | 29.4 | 38.8 | 59.1 | 74.4 | 130.3 | 132.5 | 81.6 | 96.1 | 66.5 | 90.9 | 51.9 | 82.8 | — | — | — | — | — | — | — | |
| M2Trans.Ensemble=true, VL pretraining=false2026.05 | 39.7 | 72.8 | 29.4 | 39 | 59.2 | 74.8 | 129.3 | 132.1 | 81.6 | 96 | 66.4 | 90.8 | 51.8 | 82.7 | — | — | — | — | — | — | — | |
| AoANet2021.01 | 39.4 | 71.2 | 29.1 | 38.5 | 58.9 | 74.5 | 126.9 | 129.6 | 81 | 95 | 65.8 | 89.6 | 51.4 | 81.3 | — | — | — | — | — | — | — | |
| AoANetsplit=official test, server=online test server2022.06 | 39.4 | 71.2 | 29.1 | 38.5 | 58.9 | 74.5 | 126.9 | 129.6 | 81 | 95 | 65.8 | 89.6 | 51.4 | 81.3 | — | — | — | — | — | — | — | |
| AoAEnsemble=true, VL pretraining=false2022.07 | 39.4 | 71.2 | 29.1 | 38.5 | 58.9 | 74.5 | 126.9 | 129.6 | 81 | 95 | 65.8 | 89.6 | 51.4 | 81.3 | — | — | — | — | — | — | — | |
| AoANetModel Type=Ensemble2022.12 | 39.4 | 71.2 | 29.1 | 38.5 | 58.9 | 74.5 | 126.9 | 129.6 | 81 | 95 | 65.8 | 89.6 | 51.4 | 81.3 | — | — | — | — | — | — | — | |
| AoANetEnsemble=4 models2023.05 | 39.4 | 71.2 | 29.1 | 38.5 | 58.9 | 74.5 | 126.9 | 129.6 | 81 | 95 | 65.8 | 89.6 | 51.4 | 81.3 | — | — | — | — | — | — | — | |
| AoANet2022.01 | 39.4 | 71.2 | 29.1 | 38.5 | 58.9 | 74.5 | 126.9 | 129.6 | 81 | 95 | 65.8 | 89.6 | 51.4 | 81.3 | — | — | — | — | — | — | — | |
| AoAEnsemble=true, VL pretraining=false2026.05 | 39.4 | 71.2 | 29.1 | 38.5 | 58.9 | 74.5 | 126.9 | 129.6 | 81 | 95 | 65.8 | 89.6 | 51.4 | 81.3 | — | — | — | — | — | — | — | |
| HIPEnsemble=false, VL pretraining=false2022.07 | 39.3 | 71 | 28.8 | 38.1 | 59 | 74.1 | 127.9 | 130.2 | 81.6 | 95.9 | 66.2 | 90.4 | 51.5 | 81.6 | — | — | — | — | — | — | — | |
| GCN-LSTM+HIP2022.01 | 39.3 | 71 | 28.8 | 38.1 | 59 | 74.1 | 127.9 | 130.2 | 81.6 | 95.9 | 66.2 | 90.4 | 51.5 | 81.6 | — | — | — | — | — | — | — | |
| Dual GlobalEnsemble=false, VL pretraining=false2022.07 | 39.1 | 71.2 | 28.9 | 38.4 | 58.9 | 74.4 | 126.3 | 129.2 | 80.8 | 95.1 | 65.6 | 81.3 | 51.1 | 81.3 | — | — | — | — | — | — | — | |
| APNsplit=official test, server=online test server2022.06 | 38.9 | 70.2 | 28.8 | 38 | 58.7 | 73.7 | 126.3 | 127.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| APNModel Type=Ensemble2022.12 | 38.9 | 70.2 | 28.8 | 38 | 58.7 | 73.7 | 126.3 | 127.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ETA2022.01 | 38.9 | 70.2 | 28.6 | 38 | 58.6 | 73.9 | 122.1 | 124.4 | 81.2 | 95 | 65.5 | 89 | 50.9 | 80.4 | — | — | — | — | — | — | — | |
| GCN-LSTMsplit=official test, server=online test server2022.06 | 38.7 | 69.7 | 28.5 | 37.6 | 58.5 | 73.4 | 125.3 | 126.5 | 80.8 | 95.2 | 65.5 | 89.3 | 50.8 | 80.3 | — | — | — | — | — | — | — | |
| GCN-LSTMEnsemble=true, VL pretraining=false2022.07 | 38.7 | 69.7 | 28.5 | 37.6 | 58.5 | 73.4 | 125.3 | 126.5 | 80.8 | 95.2 | 65.5 | 89.3 | 50.8 | 80.3 | — | — | — | — | — | — | — | |
| GCN-LSTMModel Type=Ensemble2022.12 | 38.7 | 69.7 | 28.5 | 37.6 | 58.5 | 73.4 | 125.3 | 126.5 | 80.8 | 95.2 | 65.5 | 89.3 | 50.8 | 80.3 | — | — | — | — | — | — | — | |
| GCN-LSTM2022.01 | 38.7 | 69.7 | 28.5 | 37.6 | 58.5 | 73.4 | 125.3 | 126.5 | 80.8 | 95.9 | 65.5 | 89.3 | 50.8 | 80.3 | — | — | — | — | — | — | — | |
| GCN-LSTMEnsemble=true, VL pretraining=false2026.05 | 38.7 | 69.7 | 28.5 | 37.6 | 58.5 | 73.4 | 125.3 | 126.5 | 80.8 | 95.2 | 65.5 | 89.3 | 50.8 | 80.3 | — | — | — | — | — | — | — | |
| SGAEsplit=official test, server=online test server2022.06 | 38.5 | 69.7 | 28.2 | 37.2 | 58.6 | 73.6 | 123.8 | 126.5 | 81 | 95.3 | 65.6 | 89.5 | 50.7 | 80.4 | — | — | — | — | — | — | — | |
| SGAEEnsemble=true, VL pretraining=false2022.07 | 38.5 | 69.7 | 28.2 | 37.2 | 58.6 | 73.6 | 123.8 | 126.5 | 81 | 95.3 | 65.6 | 89.5 | 50.7 | 80.4 | — | — | — | — | — | — | — | |
| SGAE2022.01 | 38.5 | 69.7 | 28.2 | 37.2 | 58.6 | 73.6 | 123.8 | 126.5 | 81 | 95.3 | 65.6 | 89.5 | 50.7 | 80.4 | — | — | — | — | — | — | — | |
| SGAEEnsemble=true, VL pretraining=false2026.05 | 38.5 | 69.7 | 28.2 | 37.2 | 58.6 | 73.6 | 123.8 | 126.5 | 81 | 95.3 | 65.6 | 89.5 | 50.7 | 80.4 | — | — | — | — | — | — | — | |
| AoANet+VCVC features=true, backbone=AoANet2020.02 | 38.4 | 69.9 | 28.8 | 38 | 58.6 | 73.8 | 125.5 | 128.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SCD-NetModel Type=Single2022.12 | 38.1 | 69.4 | 29 | 38.2 | 58.5 | 73.5 | 126.2 | 129.2 | 80.2 | 95.1 | 64.9 | 89.3 | 50.1 | 80.1 | — | — | — | — | — | — | — | |
| RFNetModel Type=Ensemble2022.12 | 38 | 69.2 | 28.2 | 37.2 | 58.2 | 73.1 | 122.9 | 125.1 | 80.4 | 95 | 64.9 | 89.3 | 50.1 | 80.1 | — | — | — | — | — | — | — | |
| RFNet2022.01 | 38 | 69.2 | 28.2 | 37.2 | 58.2 | 73.1 | 122.9 | 125.1 | 80.4 | 95 | 64.9 | 89.3 | 50.1 | 80.1 | — | — | — | — | — | — | — | |
| CNMVC features=false2020.02 | 37.9 | 68.4 | 28.1 | 36.9 | 58.3 | 72.9 | 123 | 125.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CAVPModel Type=Single2022.12 | 37.9 | 69 | 28.1 | 37 | 58.2 | 73.1 | 121.6 | 123.8 | 80.1 | 94.9 | 64.7 | 88.8 | 50 | 79.7 | — | — | — | — | — | — | — | |
| SGAEVC features=false2020.02 | 37.8 | 68.7 | 28.1 | 37 | 58.2 | 73.1 | 122.7 | 125.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Up-Down+VCVC features=true, backbone=Up-Down2020.02 | 37.8 | 69.1 | 28.5 | 37.6 | 58.2 | 73.3 | 124.1 | 126.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SGAEModel Type=Single2022.12 | 37.8 | 68.7 | 28.1 | 37 | 58.2 | 73.1 | 122.7 | 125.5 | 80.6 | 95 | 65 | 88.9 | 50.1 | 79.6 | — | — | — | — | — | — | — | |
| AoANetVC features=false2020.02 | 37.3 | 68.1 | 28.3 | 37.2 | 57.9 | 72.8 | 124 | 126.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Up-DownVC features=false2020.02 | 36.9 | 68.5 | 27.6 | 36.7 | 57.1 | 72.4 | 117.9 | 120.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| BUTD2021.01 | 36.9 | 68.5 | 27.6 | 36.7 | 57.1 | 72.4 | 117.9 | 120.5 | 80.2 | 95.2 | 64.1 | 88.8 | 49.1 | 79.4 | — | — | — | — | — | — | — | |
| Up-Downsplit=official test, server=online test server2022.06 | 36.9 | 68.5 | 27.6 | 36.7 | 57.1 | 72.4 | 117.9 | 120.5 | 80.2 | 95.2 | 64.1 | 88.8 | 49.1 | 79.4 | — | — | — | — | — | — | — | |
| Up-DownEnsemble=true, VL pretraining=false2022.07 | 36.9 | 68.5 | 27.6 | 36.7 | 57.1 | 72.4 | 117.9 | 120.5 | 80.2 | 95.2 | 64.1 | 88.8 | 49.1 | 79.4 | — | — | — | — | — | — | — | |
| Up-DownModel Type=Ensemble2022.12 | 36.9 | 68.5 | 27.6 | 36.7 | 57.1 | 72.4 | 117.9 | 120.5 | 80.2 | 95.2 | 64.1 | 88.8 | 49.1 | 79.4 | — | — | — | — | — | — | — | |
| Up-DownEnsemble=4 models2023.05 | 36.9 | 68.5 | 27.6 | 36.7 | 57.1 | 72.4 | 117.9 | 120.5 | 80.2 | 95.2 | 64.1 | 88.8 | 49.1 | 79.4 | — | — | — | — | — | — | — | |
| Up-Down2022.01 | 36.9 | 68.5 | 27.6 | 36.7 | 57.1 | 72.4 | 117.9 | 120.5 | 80.2 | 95.2 | 64.1 | 88.8 | 49.1 | 79.4 | — | — | — | — | — | — | — | |
| Up-DownEnsemble=true, VL pretraining=false2026.05 | 36.9 | 68.5 | 27.6 | 36.7 | 57.1 | 72.4 | 117.9 | 120.5 | 80.2 | 95.2 | 64.1 | 88.8 | 49.1 | 79.4 | — | — | — | — | — | — | — | |
| CMALModel Type=Single2022.12 | 36.8 | 66.1 | 27.9 | 36.4 | 57.6 | 72 | 119.3 | 121.2 | 79.8 | 94.3 | 63.8 | 87.2 | 48.8 | 77.2 | — | — | — | — | — | — | — | |
| HANEnsemble=true, VL pretraining=false2022.07 | 36.5 | 66.8 | 27.4 | 36.1 | 57.3 | 71.9 | 115.2 | 118.2 | 80.4 | 94.5 | 63.8 | 87.7 | 48.8 | 78 | — | — | — | — | — | — | — | |
| HANEnsemble=true, VL pretraining=false2026.05 | 36.5 | 66.8 | 27.4 | 36.1 | 57.3 | 71.9 | 115.2 | 118.2 | 80.4 | 94.5 | 63.8 | 87.7 | 48.8 | 78 | — | — | — | — | — | — | — | |
| SCSTEnsemble=true, VL pretraining=false2022.07 | 35.2 | 64.5 | 27 | 35.5 | 56.3 | 70.7 | 114.7 | 116.7 | 78.1 | 93.7 | 61.9 | 86 | 47 | 75.9 | — | — | — | — | — | — | — | |
| SCSTModel Type=Ensemble2022.12 | 35.2 | 64.5 | 27 | 35.5 | 56.3 | 70.7 | 114.7 | 116.7 | 78.1 | 93.7 | 61.9 | 86 | 47 | 75.9 | — | — | — | — | — | — | — | |
| SCSTEnsemble=4 models2023.05 | 35.2 | 64.5 | 27 | 35.5 | 56.3 | 70.7 | 114.7 | 116.7 | 78.1 | 93.7 | 61.9 | 86 | 47 | 75.9 | — | — | — | — | — | — | — | |
| SCST2022.01 | 35.2 | 64.5 | 27 | 35.5 | 56.3 | 70.7 | 114.7 | 116.7 | 78.1 | 93.7 | 61.9 | 86 | 47 | 75.9 | — | — | — | — | — | — | — | |
| SCSTEnsemble=true, VL pretraining=false2026.05 | 35.2 | 64.5 | 27 | 35.5 | 56.3 | 70.7 | 114.7 | 116.7 | 78.1 | 93.7 | 61.9 | 86 | 47 | 75.9 | — | — | — | — | — | — | — | |
| AoANetreference captions=402026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 95 | 89.6 | 81.3 | 71.2 | 38.5 | 74.5 | 129.6 | |
| DLCTreference captions=402026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 96.2 | 91 | 83 | 73.2 | 39.1 | 74.8 | 133.4 | |
| DSCTreference captions=402026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 97.3 | 91.5 | 83.8 | 73.9 | 39.8 | 75.3 | 136.7 | |
| ETAIreference captions=402026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 95 | 89 | 80.4 | 70.2 | 38 | 73.9 | 124.4 | |
| GCN-LSTMreference captions=402026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 95.9 | 89.3 | 80.3 | 69.7 | 37.6 | 73.4 | 126.5 | |
| M2 Transformerreference captions=402026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 96 | 90.8 | 82.7 | 72.8 | 39 | 74.8 | 132.1 | |
| RFNetreference captions=402026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 95 | 89.3 | 80.1 | 69.2 | 37.2 | 74.1 | 125.1 | |
| RSTNetreference captions=402026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 96.2 | 90.9 | 82.7 | 72.5 | 38.7 | 74.2 | 132.4 | |
| S2-Transformerreference captions=402026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 96.4 | 91.3 | 83.1 | 73.1 | 39.2 | 74.7 | 134.5 | |
| SCSTreference captions=402026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 93.7 | 86 | 75.9 | 64.5 | 35.5 | 70.7 | 116.7 | |
| SGAEreference captions=402026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 95.3 | 89.5 | 80.4 | 69.7 | 37.2 | 73.6 | 126.5 | |
| Up-Downreference captions=402026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 95.2 | 88.8 | 79.4 | 68.5 | 36.7 | 72.4 | 120.5 | |
| X-Transformerreference captions=402026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 95.4 | 90 | 81.7 | 71.8 | 39 | 74.9 | 131.4 |