Cross-modal retrieval on MSCOCO 1K
91.6Mean Recall (ja)MURAL-LARGE
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MURAL-LARGEEvaluation Protocol=Fine-tuned, Data=AT+MBT, Type=Dual Encoder2021.09 | 91.6 | 92.3 | |
| MURAL-BASEEvaluation Protocol=Fine-tuned, Data=AT+MBT, Type=Dual Encoder2021.09 | 88.4 | 88.6 | |
| M3PEvaluation Protocol=Fine-tuned, Data=CC3m+Wiki, Type=Cross Encoder2021.09 | 87.9 | 88.7 | |
| UC2Evaluation Protocol=Fine-tuned, Data=TrTrain(CC3m), Type=Cross Encoder2021.09 | 87.5 | 88.1 | |
| ALIGN-BASEEvaluation Protocol=Fine-tuned, Data=TrTrain(AT-en), Type=Dual Encoder2021.09 | 87.5 | 89 | |
| MURAL-BASEEvaluation Protocol=Fine-tuned, Data=TrTrain(CC12m)+EOBT, Type=Dual Encoder2021.09 | 87.4 | 89.4 | |
| ALIGN-BASEEvaluation Protocol=Fine-tuned, Data=AT, Type=Dual Encoder2021.09 | 86.7 | 89.2 | |
| MURAL-LARGEEvaluation Protocol=Zero-shot, Data=AT+MBT, Type=Dual Encoder2021.09 | 81.3 | 84.4 | |
| SMALREvaluation Protocol=Fine-tuned, Data=no pretraining, Type=Dual Encoder2021.09 | 77.5 | 81.5 | |
| MURAL-BASEEvaluation Protocol=Zero-shot, Data=AT+MBT, Type=Dual Encoder2021.09 | 73.4 | 79.2 | |
| MURAL-BASEEvaluation Protocol=Zero-shot, Data=TrTrain(CC12m)+EOBT, Type=Dual Encoder2021.09 | 72.5 | 78.1 | |
| ALIGN-BASE-ENEvaluation Protocol=Zero-shot, Data=AT-en-translate-test, Type=Dual Encoder2021.09 | 71.5 | 80 | |
| ALIGN-BASEEvaluation Protocol=Zero-shot, Data=AT, Type=Dual Encoder2021.09 | 70.9 | 79.5 | |
| ALIGN-L2Evaluation Protocol=Zero-shot, Data=AT-en, Type=Dual Encoder2021.09 | 70.9 | — | |
| ALIGN-BASEEvaluation Protocol=Zero-shot, Data=TrTrain(AT-en), Type=Dual Encoder2021.09 | 70.6 | 77.1 | |
| M3PEvaluation Protocol=Zero-shot, Data=CC3m+Wiki, Type=Cross Encoder2021.09 | 33.3 | 63.1 |