VGGNet based DNN for recognition skeletonized character
Skeletonize | Deep-text | CRAFT
This model is for performance assistance in cases where the recognition rate is low in the existing OCR performance, and consists of Skeletonize and VGG.
17.08.2020 : Initial update
Now attending...
Total dataset is based by their dataset, extract if that image's deep-text-recognition conf score are over 60%