Training one common model on two different type of images

clovaai / donut

Official Implementation of OCR-free Document Understanding Transformer (Donut) and Synthetic Document Generator (SynthDoG), ECCV 2022

https://arxiv.org/abs/2111.15664

MIT License

5.53k stars 444 forks source link

Training one common model on two different type of images #220

Open NavneetSingh20 opened 1 year ago

NavneetSingh20 commented 1 year ago

I have been trying to train a common model that can extract data from two different type of images. I am getting all other details but it is not extracting the ID numbers of these two type of images. Can anyone suggest something to improve the results.