Semantic Image-Text Similarity on CxC
74.58Avg ScoreCC:DAE
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| CC:DAEEvaluation Protocol=Zero-shot2024.02 | 74.58 | — | 67.71 | |
| Meaning as TrajectoriesEvaluation Protocol=Zero-shot2024.02 | 74.55 | — | 67.63 | |
| CLIP-ViT B/16Backbone=ViT-B/16, Evaluation Protocol=Zero-shot2024.02 | 68.34 | — | 64.6 | |
| ALIGN-L2Fine-tuning Data=MSCOCO2021.09 | 67.6 | 1.2 | — | |
| MURAL-LARGEFine-tuning Data=Multi30K and MSCOCO2021.09 | 67.1 | 1.3 | — | |
| MURAL-BASEFine-tuning Data=Multi30K and MSCOCO2021.09 | 64 | 1.2 | — | |
| ALIGN-BASEFine-tuning Data=Multi30K and MSCOCO2021.09 | 63.7 | 1.3 | — | |
| DE-T2T+I2TFine-tuning Data=MSCOCO2021.09 | 61.9 | 1.3 | — |