Multi-modal In-Context Learning Makes an Ego-evolving Scene Text Recognizer

Benchmark Model Rank Results
scene-text-recognition-on-wordartE²STR-ICL#1Accuracy: 86.17
scene-text-recognition-on-cute80E²STR-ICL#5Accuracy: 99.31
scene-text-recognition-on-coco-textE²STR-ICL#61:1 Accuracy: 78.38
scene-text-recognition-on-svtE²STR-ICL#6Accuracy: 98.61
scene-text-recognition-on-wostE²STR-ICL#61:1 Accuracy: 86.59
scene-text-recognition-on-iiit5kE²STR-ICL#9Accuracy: 99.23
scene-text-recognition-on-svtpE²STR-ICL#10Accuracy: 96.74