Vision Transformer for Fast and Efficient Scene Text Recognition

Benchmark Model Rank Results
scene-text-recognition-on-icdar-2003ViTSTR#9Accuracy: 94.3
scene-text-recognition-on-icdar2013-857-imagesViTSTR-Base + RandAugment (extra data)#18Accuracy: 93.2
scene-text-recognition-on-icdar2015-1-811-imagesViTSTR-Base + RandAugment (extra data)#19Accuracy: 78.5
scene-text-recognition-on-icdar2015ViTSTR#31Accuracy: 72.6
scene-text-recognition-on-icdar2013ViTSTR#37Accuracy: 92.4
scene-text-recognition-on-svtpViTSTR-Base + RandAugment (extra data)#39Accuracy: 81.8
scene-text-recognition-on-cute80ViTSTR-Base + RandAugment (extra data)#45Accuracy: 81.3
scene-text-recognition-on-iiit5kViTSTR-Base + RandAugment (extra data)#46Accuracy: 88.4
scene-text-recognition-on-svtViTSTR#51Accuracy: 87.7