Scene Text Recognition on SVTP Cleaned (test)
97.4Word AccuracyCLIP4STR-B
Evaluation Results
| Method | Links | |
|---|---|---|
| CLIP4STR-BTraining Data=3.3M real samples, Backbone=ViT-B2023.05 | 97.4 | |
| CLIP4STR-LTraining Data=3.3M real samples, Backbone=ViT-L2023.05 | 97.4 | |
| MPSTRTraining Data=3.3M real samples2023.05 | 96.1 | |
| PARSeqTraining Data=3.3M real samples2023.05 | 95.7 | |
| ABINetTraining Data=3.3M real samples2023.05 | 94.7 |