Text-to-Image Retrieval on Stock 10K (test)
33.83R@1Multimodal Contrastive Training
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Multimodal Contrastive TrainingBackbone=Frozen visual backbone2021.04 | 33.83 | 70.76 | 3 | |
| Stock-supTraining=Supervised on tagging task2021.04 | 30.64 | 68.18 | 4 |